L-DACS2 System Definition Proposal: Deliverable D2

Transcription

1 EUROPEAN ORGANISATION FOR THE SAFETY OF AIR NAVIGATION EUROCONTROL L-DACS2 System Definition Proposal: Deliverable D2 Edition Number : v1.0 Edition Date : 11 th May 2009 Status : Draft Intended for : General Public EUROPEAN AIR TRAFFIC MANAGEMENT

2 DOCUMENT CHARACTERISTICS TITLE LDACS2 System Definition Proposal: Deliverable D2 Publications Reference: ISBN Number: Document Identifier Edition Number: 1.0 Edition Date: 11/05/2009 Abstract Keywords LDACS2 Data link L band FCI AMACS GSM LDL Authors Contact(s) Person Tel Unit Nikos Fistas CND/CoE/CNS/COM STATUS, AUDIENCE AND ACCESSIBILITY Status Intended for Accessible via Working Draft General Public Intranet Draft EATM Stakeholders Extranet Proposed Issue Restricted Audience Internet ( Released Issue Electronic copies of this document can be downloaded from LBANDLIB.html Edition: 1.0 Draft Page 2

3 DOCUMENT APPROVAL The following table identifies all management authorities who have successively approved the present issue of this document. AUTHORITY NAME AND SIGNATURE DATE Edition: 1.0 Draft Page 3

4 DOCUMENT CHANGE RECORD The following table records the complete history of the successive editions of the present document. EDITION NUMBER EDITION DATE REASON FOR CHANGE PAGES AFFECTED Publications EUROCONTROL Headquarters 96 Rue de la Fusée B-1130 BRUSSELS Tel: +32 (0) Fax: +32 (0) Edition: 1.0 Draft Page 4

5 Contents CHAPTER 1 Introduction Background Objective and scope of the document Project Team L-DACS2 design L-DACS2 overall description Outline of the specification...19 CHAPTER 2 Functional Architecture Introduction Infrastructure Communication link...21 CHAPTER 3 Air interface (physical layer) Introduction Transmitter/receiver frequency control Digital reception by the receiver Digital transmission Modulation Physical layer Standards Modulation scheme Modulation rate Time/amplitude profile of L-DACS2 transmission Ambiguity resolution and data transmission Receiver - transmitter turnaround time Frequency change during transmission Air interface Radio frequencies Channel bandwidth Polarization Ground Frequency stability Aircraft Frequency stability Spurious emissions Error phase specification Broad band noise Connection management Transmission Channel...29 Edition: 1.0 Draft Page 5

6 Broadcast Air-to-air point-to-point communication System performance Link Budget Pulse blanking Protection of L-DACS2 transmissions Protection of adjacent systems transmissions Power control Received signals ground Received signals aircraft Single Antenna Interference Cancellation Interference immunity Transmitting function Adjacent channel emissions Receiving function Specified error rate Reference sensitivity level Adjacent band immunity performance and co-channel immunity performance Out-of-band immunity performance Interference immunity performance FEC mechanism Inner code: convolutive punctured code Interleaver Outer code: Reed Solomon code...39 CHAPTER 4 MAC sublayer Introduction Provision MAC Layer for point-to-point and broadcast communication Framing Synchronization Specified time reference Primary time synchronization mode Secondary synchronization mode Derived synchronization mode Synchronisation of slots within a frame Reversion Burst format Burst composition Bursts occupying multiple slots Ground station bursts Framing structure Frame Structure...46 Edition: 1.0 Draft Page 6

7 4.5.2 Login and response Frame section length alteration Framing structure default parameter values and ranges Adaptive burst alignment Burst transmission in relation to slot and frame structure UP1 and UP2 sections LoG2 insertion section CoS1 and CoS2 sections CoS1 section CoS2 section Summary of burst guard times Quality of Service (QoS) management Efficiency L-DACS2 QoS management Parameters Priority information End-to-end QoS management Processing Power Control management MAC layer for Broadcast Service Frame Synchronisation Burst format Frame structure MAC layer for Air-to-Air Point-to-Point Service Frame Synchronisation Burst format Frame structure...59 CHAPTER 5 Data link sublayer (DLS) Introduction Transmission procedure Uplink transmission procedures Downlink transmission procedures Reception procedure Segmentation Reserved access protocol specification Random access protocol specification for transmission in CoS Random access parameters Parameter p1 (Persistence CoS2) Counter VS3 (maximum number of access attempts) Random access procedures Random access protocol specification for transmission in LoG Edition: 1.0 Draft Page 7

8 5.7.1 Random access parameters Parameter p2 (Persistence LoG2) Counter VS4 (maximum number of access attempts) Random access procedures...65 CHAPTER 6 L-DACS2 Link Management layer (LML)...66 ANNEX 1 ANNEX Introduction Login mechanism Hand-over mechanism Controlled hand-over Air-initiated controlled hand-over Ground-requested air-initiated controlled hand-over Uncontrolled hand-over...77 Example of burst transmission...78 A1.1 Burst transmission in UP1 or UP Burst transmission in a single slot in CoS Burst transmission in CoS1 slot Burst transmission in a LoG2 slot...84 Message structure...87 A2.1 Message type codes...87 A2.2 Message type codes...87 A2.3 Priority field...88 A2.4 Messages...88 ANNEX 3 System operations...99 A3.1 Downlink...99 A3.1.1 A/C login...99 A3.1.2 Aircraft has data to send A3.1.3 Aircraft has no data to send A3.1.4 CoS2 random access A3.1.5 Aircraft-initiated cell exit A3.1.6 GS request for aircraft cell exit with no recommendation A3.1.7 GS request for aircraft cell exit with recommendation A3.2 Uplink A3.2.1 GS has data to send A Data size is 2,048 octets, if transmitted in UP A Data size is >2,048 octets A3.3 Acknowledgement messages ANNEX 4 Coding and interleaving A4.1 GMSK and convolutional coding: rate and expected performances A4.2 Considerations regarding practical C/N A4.3 Accounting for burst interference: interleaving and RS coding A4.4 Equalization ANNEX 5 Impact of intra-system interference Edition: 1.0 Draft Page 8

9 A5.1 Intra system interference robustness A5.2 Adjacent channel frequency guard time A5.3 Co-channel spacing guard band A5.4 Multiple channels operating in one cell A5.4.1 Frequency Band A5.4.2 Spacing separation between stations Edition: 1.0 Draft Page 9

10 List of Figures Figure 1: Selection process of L-DACS system...15 Figure 2: Proposed Approach...16 Figure 3: Time/amplitude profile of LDACS2 transmission...24 Figure 4 L-DACS2 Link Budget...31 Figure 5 Adjacent band and Co-channel band reference interference ratio...37 Figure 6 : L-DACS/2 frame...41 Figure 7: L-DACS/2 burst structure for downlink bursts...44 Figure 8: L-DACS/2 merged slot structure for downlink bursts...45 Figure 9: L-DACS/2 merged slot structure structure for downlink bursts...45 Figure 10: L-DACS/2 merged slot structure for uplink bursts...46 Figure 11: L-DACS2 frame structure (point-to-point)...47 Figure 12: Successful air-initiated controlled hand-over...69 Figure 13: Air-initiated controlled handover: retransmit cell login...69 Figure 14: Air-initiated controlled handover: retransmit GS_ALLOC...70 Figure 15: Air-initiated controlled handover: retransmit AC_CELL_EXIT...70 Figure 16: Air-initiated controlled handover: retransmit GS_EXIT_ACK...71 Figure 17: Successful ground-requested air-initiated controlled hand-over...73 Figure 18: Ground-requested air-initiated controlled hand-over: GS_CELL_EXIT retransmit73 Figure 19: Ground-requested air-initiated controlled hand-over: AC_EXIT_ACK retransmit.74 Figure 20: Ground-requested air-initiated controlled hand-over: GS_FRAME not received..74 Figure 21: Ground-requested air-initiated controlled hand-over: unsuccessful login...75 Figure 22: Successful ground-requested air-initiated controlled hand-over with recommendation...75 Figure 23: Ground-requested air-initiated controlled hand-over with recommendation: alternative ground station...76 Figure 24: Ground-requested air-initiated controlled hand-over with recommendation: no successful logins...76 Figure 25: Uncontrolled handover...77 Figure 26: Illustration of burst format for a typical burst in UP1 or UP Figure 27: Illustration of burst format for a single slot burst in CoS Figure 28: Illustration of burst format for a burst in a CoS1 slot...84 Figure 29: Illustration of burst format for a burst in a LoG2 slot...86 Figure 30: GMSK theoretical performance Figure 31: Convolutional code performance Figure 32: Convolutional code (5,7), constraint length Edition: 1.0 Draft Page 10

11 Figure 33: Convolutional code (133,171), constraint length Figure 34 Separation distances between aircraft and nearest ground station Edition: 1.0 Draft Page 11

12 List of Tables Table 1 Total channel bandwidth for various cell sizes...26 Table 2 Latencies for TMA Large scenario (52 aircraft)...27 Table 3 Latencies for En route Medium scenario (62 aircraft)...27 Table 4 Latencies for En route large scenario (102 aircraft per channel)...27 Table 5 Throughput achieved...27 Table 6 Spectrum mask (db)...28 Table 7 Spurious domain emissions ITU regulation...29 Table 8 Effect of imposing burst length limits in the Reverse Link...34 Table 9 FLIPINT and COTRAC messages sizes (downlink)...34 Table 10: Parameter definition for framing parameters...41 Table 11: Parameter definition for framing parameters...47 Table 12: Parameter definition for S1 parameter...48 Table 13: Parameter definition for framing parameters...50 Table 14: Parameter definition for timing advance parameter T Table 15: Allocation of T5 parameter to transmission delay...52 Table 16: End guard period for UP1 and UP2 burst as a function of the number of slots occupied by the burst...52 Table 17: Parameter definition for maximum burst length in slots in CoS Table 18: Guard period for CoS2 slot as a function of slot length...55 Table 19: Summary of burst guard periods per frame section...56 Table 20: Summary of burst guard periods at the start of frame sections...56 Table 21: Definition of parameter Priority Q Table 22: Mapping between message category, ATN priority and Q22 priority classification57 Table 23: Random access parameters for CoS Table 24: Random access parameters for LoG Table 25: Available payload for user data in burst occupying two slots in UP1 or UP Table 26: Available payload for user data in a single slot in CoS Table 27: Available payload for user data in a CoS1 slot...83 Table 28: Available payload for user data in a LoG2 slot...85 Table 29 Link budget including interference contributions Table 30 Adjacent channel frequency guard time Table 31 Co-channel spacing guard band Table 32 Multiple channels operating in one cell scenario Edition: 1.0 Draft Page 12

13 EXECUTIVE SUMMARY This document is the first of the deliverables of the study funded by EUROCONTROL to provide the specifications for the L-DACS2 system. L-DACS2 is one of the two candidate systems identified in ICAO and the SESAR Definition Phase to support the future aeronautical communications infrastructure (FCI) in the continental enroute and TMA environments. L-DCAS2 is considered for operation in the L band, and has been developed from material for AMACS, UAT, DME and VDL Mode 4. GSM technical elements have also been used as basic technical background. This document will be distributed for external review by the interested parties including the SJU WP9 and WP15 partners. Following the review the document will be updated to consider the received comments and the new document (Deliverable D2) will provide the basis for the development of detailed Tx and Rx prototype equipment specifications. These prototypes will be used to demonstrate the spectrum compatibility of the candidate systems with the existing systems operating in the L band and the suitability of its performance in the presence of interference from the existing systems. These activities will facilitate the eventual selection of one system (L- DACS) for the FCI The deliverables of this study will be an input to the SJU relevant projects (in particular P15.2.4). Edition: 1.0 Draft Page 13

14 CHAPTER 1 Introduction 1.1 Background In support of the future SJU activities, EUROCONTROL is leading the investigations for the future communications infrastructure (FCI) which is required to support the future aeronautical communications. One of the key recommendations concerning the development of a new data link, issued in the AP17 Final Report and in SESAR Definition Phase Deliverables, is to finalise the selection of a an L-Band data link (LDACS) that supports the continental airspace environment. The development of the L-band data link is identified in the development activities for the SESAR Implementation Package 3 (IP3) in the post 2020 timeframe. Therefore, the outcome of this activity will be used as input to the SESAR JU activities. Under the AP17 activities, various candidate technologies were considered and evaluated to operate in the L band and supporting the requirements. However, it was found that none of the considered technologies could be fully recommended primarily due to concerns about the operational compatibility (spectrum interference) with existing systems in the L band. Nevertheless, the assessment of the candidate technologies led to the identification of desirable technology features that could be used as a basis for the development of an L- band data link solution that would be spectrally compatible. Considering these features and the most promising candidates, two options for the L-band Digital Aeronautical Communication System (LDACS) were identified. These options need further consideration before final selection of a single data link technology. The first option for LDACS is a frequency division duplex (FDD) configuration utilizing OFDM modulation techniques, reservation based access control and advanced network protocols. This solution is a derivative of the B-AMC and TIA-902 (P34) technologies. The second LDACS option is a time division duplex (TDD) configuration utilizing a binary modulation derivative of the implemented UAT system (CPFSK family) and of existing commercial (e.g. GSM) systems and custom protocols for lower layers providing high qualityof-service management capability. This solution is a derivative of the LDL and AMACS technologies. AP17 and SESAR proposed follow on activities in order to further specify the proposed Edition: 1.0 Draft Page 14

15 LDACS option, validate their performance, aiming at a final decision (single technology recommendation for L-Band) by Based on the information given above, in order to complete the selection of the LDACS, it is required to: Develop detailed specifications for LDACS1 and LDACS2 Develop and test LDACS1 and LDACS2 prototypes Assess the overall performance of LDACS1 and LDACS2 systems. The figure bellow explains the complete process applied to select the final L-DACS system. The activity conducted under this contract (LDACS2 study) only concern the shaded box and the LDACS2 system. A separate Eurocontrol contract (LDACS1 study) is addressing in a similar manner the LDACS1 system. Figure 1: Selection process of L-DACS system 1.2 Objective and scope of the document The L-DACS2 Study objective is to support the realisation of the recommendation for the selection of the L band data link which is expected to be progressed in the frame of the SJU WP15 activities. Under these activities, it is expected that industry will develop LDACS1/2 prototypes for testing in order to evaluate the spectrum compatibility of LDACS1/2 with the existing users of the L band and the overall capabilities of the LDACS1/2 system and eventually facilitate the LDACS system selection. In this context, the LDACS2 study will provide input to SJU activities (Project ) by developing initial specifications for the L-DACS2 system in sufficient detail so as to enable the subsequent development (outside the scope of this project) of L-DACS2 prototype(s) for testing (aiming to confirm the spectrum compatibility of L-DACS2 with the existing users of the L band and to demonstrate and validate the capabilities of the system). The LDACS2 system specification trade-off aims to achieve the following high level Edition: 1.0 Draft Page 15

16 objectives: L-DACS2 System Definition proposal: Deliverable D2 L-DACS2 can operate in the L band without interfering with the existing users of the band. The performance of the L-DACS2 is meeting the expected requirements in the expected operational environment. L-DACS2 development is facilitated and expedited through the choice of appropriate components and/or mature standards. The proposed specifications should be sufficiently complete and unambiguous to be able to initiate the system design validation that will lead (with an iterative refinement cycle if required) to proposals for international standards (ICAO, EUROCAE/RTCA). There will be two main outputs from the L-DACS2 study. The first one will be the L-DACS2 system detailed characteristics. For this there will be an initial version for external review and comments (Deliverable D1) and the final one taking into account the external comments (Deliverable D2). The second outcome will be the detailed design specifications (Deliverable D3) for the L-DACS2 Receiver and Transmitter prototype equipment to be used in the testing. This document represents Deliverable D2. For the external review, EUROCONTROL is soliciting comments from the SJU WP15 and WP9 partners, as well as from the US and all other interested parties. Based on the outcome of the AP17 investigations, L-DACS2 draws from the features of the AMCAS and LDL systems that have been considered in the AP17 investigations. In order to develop the L-DACS2 system specifications, the L-DACS2 study partners have used the AMACS system design as the baseline for the L-DACS2 system specification considering other features form LDL (UAT) and GSM as appropriately. For this the AMACS Description v1.0 (see bellow section 1.5.1) was analysed by the LDACS2 study partners, identifying constraints, and defining a solution during an internal workshop with industry participation. These solutions were further analysed and updated after the workshop. The product of this activity was an L-DACS2 Detailed Design, as illustrated below: Figure 2: Proposed Approach Edition: 1.0 Draft Page 16

17 1.3 Project Team L-DACS2 System Definition proposal: Deliverable D2 The LDCAS2 study has been performed by the following team: EGIS AVIA has gained an important experience in international and multipartner project management. EGIS AVIA was also involved in several communication system projects including a strong involvement in E-TDMA definition as well as AMACS concept design. Helios have supported the development of the future communications activities over a number of years, by providing impartial technical advice from its pool of communications expertise and supporting EUROCONTROL develop and maintain the future deployment strategy for aeronautical mobile communications through the Action Plan 17 activities. Helios has also undertaken considerable independent technical evaluation work by simulating and validating the leading candidate technologies including AMACS B-AMC and P34. SWEDAVIA, being a subsidiary company to LFV, has been deeply involved in development and standardisation of VDL Mode 4. LFV initiated the XDL activities, i.e. VDL Mode 4 implemented in the L-band, which is one of the foundations of AMACS. LFV has also been very active in the ICAO ACP activities regarding Future Communication System. DSNA/DTI has initiated the E-TDMA activities in 1998, in cooperation with EGIS AVIA. This initial work has been one of the foundations of the AMACS concept. DTI has also been very active in the ICAO ACP activities regarding the Future Communication System. Telerad is France's supplier of Ground to Air Aeronautical Telecommunication Systems for both Civil and Military areas. Telerad equipment is in operation in more than 60 different Countries. CNSS has gained experience in AIS and VDL Mode 4, acquired through its contracts with LFV, Malmo Aviation, Skyways and Arlanda airport. Selex is an important VHF/UHF Base Stations provider for Ground-Air- Ground communications, with a strong presence in Europe, Asia and South America. Rockwell-Collins has developed an airborne S-TDMA (VDL4) radio to support the NUP II+ project and is currently developing a High Capacity Data Radio in partnership with Thales. Rockwell Collins France has a large panel of skills in digital signal processing and communication protocols, as well as a strong experience in rapid design and prototyping. RCF are involved in numerous studies and development of aeronautical communication solutions with EUROCONTROL for several years. AVTECH has participated in standardization committees for VDL Mode 4 as a solution for ADS-B both on behalf of ICAO and within RTCA and EUROCAE. AVTECH has also participated in projects for On-Board GSM technology. Edition: 1.0 Draft Page 17

18 1.4 L-DACS2 design L-DACS2 System Definition proposal: Deliverable D2 A general description of the L-DACS2 system is proposed bellow following by the outlines of the specification L-DACS2 overall description The L-DACS2 system is a multipurpose communication system based on the AMACS concept and architecture. Its key design drivers are geared towards achieving a balance between flexibility, scalability, determinism and capitalising on available standards. The overall goal is to provide a cost-effective solution that provides the capabilities that meet the demands of the future aeronautical communications environment, while providing and viable and feasible solution in the intended time frame. The air interface is based on modulation scheme derived from GSM and UAT. This is complimented by an error recovery system based upon a strong data coding to ensure the robustness of the link for the highest Quality of Service (QoS) in terms of latency. The pool of radio frequencies intended for L-DACS2 is in the lower L-band, from MHz. The lowest assignable frequency shall be MHz (for spectrum compatibility) and the highest assignable frequency shall be 975 MHz. The L-DACS2 bandwidth is set at 200 khz offering a gross data rate of kbit/s per channel. GMSK modulation scheme with a modulation index h of 0.5 and a BT product of 0.3 has been retained in order to gain some experience from GSM standard. The L-DACS2 system has the capability to provide airground communication service in a 200 NM radius. The channel bandwidth and data rate allow the L-DACS2 system to achieve the stringent requirement in term of latency for the stringent operational scenario in term of capacity and demands. The L-DACS2 specifications also provide a link budget, a spectrum mask and the receiver immunity performance in order to have the requirements for the design of a prototype. Those physical specifications, based on the GSM and UAT standards, are in line with ITU regulation in term of spectral compatibility. The waveform is design to minimise the spectral footprint while minimising customisation of already available waveforms. The ramp up and down is defined to minimise the impact of the L-DACS2 system in the L-Band environment. Power control specifications and Pulse blanking has also been studied and guidelines are provided. Finally, a complete study on the coding scheme is conducted that lead to the implementation of a Reed Salomon plus convolutional coding to assure a corrected BER of An interleaver is also implemented in order to reduce the impact of pulsed and bursted interfering on the L-DACS2 avionics communication. The AMACS MAC layer has been updated to improve the delivery of time-critical messages. The frame length has been set to 1 second to accommodate the most stringent requirements. The L-DACS2 system is a half duplex system based on time division providing specific sections for uplink and downlink transmission. The frame consists of two uplink sections, UP1 and UP2, and two downlink sections CoS1 and CoS2, and a login section LoG2. The length of each section can be change dynamically to accommodate the operational need in one cell. The CoS1 section is composed of a number of slots which are allocated deterministically to the mobiles logged in the cell. A mobile could send specifics messages (Keep-Alive, Acknowledgement or Request to Send) in its unique CoS1 slots. A synchronisation scheme between the ground station and its mobiles based on adaptive burst alignment has been integrated in order to minimise the spectral resource loss in the guard band. The UP1 and UP2 sections are used by the ground station to communicate with the mobiles providing also synchronisation information, acknowledgment and allocated slots to the mobiles. A priority management scheme has been included to increase the efficiency of the L-DACS2 performance in term of latency for the highest QoS services. The CoS2 section is reserved for downlink reserved transmission or random transmission. The protocol applicable to each section is provided. The L-DACS2 system is designed to handle performance regimes spanning several operational scenarios from low level airport operations to TMA and En-route in high-density Edition: 1.0 Draft Page 18

19 airspace, while covering a range of demand profiles for handling of ATS and AOC services. The system has a built-in efficient air-initiated cell handover mechanism, which uses aircraft knowledge of cell locations and characteristics through on-board databases, Electronic Flight Bags (EFB) or a Common Signalling Channel (CSC). There is a specific section in the frame, (LoG2), designed for the insertion in the cell. The login mechanism and hand over are described and illustrated in this document Outline of the specification The specification presented herein focuses on elements of the L-DACS2 system design that are relevant for the subsequent development of the ICAO standard. This specification may require further iterations after completion of this ongoing EUROCONTROL activity. These are expected to be carried out with the framework of the SESAR JU development activities (WP 15). This document is structure as follows: Chapter 1 provides general overview of the L-DACS concept and an overview of the L-DACS2 main characteristics. It explains the scope of the L-DACS2 study and presents the team behind this activity. Chapter 2 provides the general architecture of the L-DACS2 system. Chapter 3 provides the physical layer specification. Chapter 4 provides the Medium Access sub layer specification. Chapter 5 provides the Data Link sub layer specification. Chapter 6 provides the Link Management Layer specification. Annex 1 provides the burst structure in each frame section. Annex 2 provides the L-DACS2 message structure. Annex 3 provides example of system operations. Annex 4 provides first analysis of the coding scheme performance. Annex 5 provides an analysis of the impact of co-channel and adjacent channel interference. Edition: 1.0 Draft Page 19

20 CHAPTER 2 Functional Architecture 2.1 Introduction L-DACS2 shall be able to simultaneously handle at least aircraft per cell in high-density airspace. L-DACS2 shall have an efficient air-initiated cell hand-over mechanism, which uses aircraft knowledge of ground station cell locations and characteristics to ensure a handover step without affecting the actual Quality of Service (QoS). L-DACS2 shall also have a ground-requested air-initiated cell hand-over mechanism for use when needed. Note. The information about cell parameters, ground station cell locations and the frequencies necessary for the aircraft to initiate and maintain contact may be available through an on-board database (uploaded/updated before the start of the flight). Note. The deployment of L-DACS2 is considered in the lower of ARNS frequency band (i.e MHz) in support of new ATM point-to-point services requiring a high QoS, giving support to SESAR or NEXTGEN future concept. The VHF band constitutes also a potential frequency band for the L-DACS2 extension. L-DACS2 is designed to be flexible and configurable, for use for point-to-point and broadcast communications. The aircraft can use L-DACS2 to communicate with aircraft (air/air point to point) as well as with the ground station (using the appropriate channels), and the ground station can selectively communicate with individual or all aircraft (multicast capability). As far as possible, in order to reduce the time to market and some validation aspects, existing GSM radio technology is proposed to be used. Broadcast services could be provided in a segregated channel using specific system configuration parameters. This service is not addressed in this document. Air-to-air data communication could be provided in other segregated channels using specific 1 Simulations have shown that a single cell can support 204 aircraft. In most cases this is possible on a single channel. For the densest scenarios with combined ATS and AOC traffic, two channels will be required with 102 aircraft served on each channel (see section Error! Reference source not found.). Edition: 1.0 Draft Page 20

21 system configuration parameters. This service is not addressed in this document. Note. L-DACS2 could support in a seamless manner AOC data communications if the necessary extra spectrum is available. 2.2 Infrastructure The ground L-DACS2 infrastructure shall comprise a number of L-DACS2 Ground Radio Stations, which are organized into clusters. Note. Typically, the Ground Radio Stations in a cluster will be geographically adjacent, providing overlaps of coverage (using different frequencies) in order to achieve the cell handover in a transparent manner ( make before break concept). Each Ground Radio Station in a cluster shall be connected to some redundant concentrator, the Ground Network Interface (GNI), which interfaces it to the transport network. Note. This interface may be via an ATN Air/Ground Router. The Air/Ground Routers supporting each cluster shall be interconnected by a ground transport network. This network shall also support Ground/Ground Routers for interconnection with end-users. Note. From this description, the ATN A/G Routers are ground-based users of the L-DACS2 sub-network service and the airborne ATN routers are mobile users of the L-DACS2 subnetwork service. 2.3 Communication link The aircraft communications system support air ground communication in point to point. The L-DACS2 system also supports broadcast air ground communication in dedicated channels. The aircraft communication system exchanges information with the station in charge of the cell in which the aircraft is located. This ground station shall provide all the communication services required to maintain the communication over the whole logged mobile and the station. Note. The L-DACS2 system could be adapted to handle future air-air point to point and broadcast services but those services are not developed in this document. This document only addresses air ground point to point services. Edition: 1.0 Draft Page 21

22 CHAPTER 3 Air interface (physical layer) 3.1 Introduction The physical layer shall provide the following functions: transmitter and receiver frequency control; digital signal reception and demodulation and decoding by the receiver; digital signal coding modulation and transmission by the transmitter; and notification services Transmitter/receiver frequency control The L-DACS2 physical layer shall set the transmitter or receiver frequency as commanded by the LME Digital reception by the receiver The receiver shall demodulate and decode input signals and forward them to the higher layers for processing Digital transmission The physical layer shall appropriately encode, modulate and transmit information received from higher layers over the RF channel. The L-DACS2 system shall make use of specific channels for air/ground point-to-point communications in a given cell. The channel shall allow air stations to have an exclusive slot per frame for regular or high-qos transmissions, with more slots available in the same frame on request. Note. The frame is fully defined in Chapter 4. Each time the word frame is used refers to the frame described in Chapter 4. Edition: 1.0 Draft Page 22

23 3.2 Modulation Physical layer L-DACS2 System Definition proposal: Deliverable D2 For L-DACS2, a robust physical layer shall be used, based on the GMSK modulation scheme used in commercial implementations like GSM. The physical layer will implement strong data coding for achieving the highest QoS in terms of latency by minimising the need for reemission of messages Standards The L-DACS2 Physical Layer uses features and characteristics of GSM, for which international standards are available ([4], [5], [6], [7], [8], [9]) Modulation scheme The modulation scheme shall be GMSK. Binary ones and binary zeros shall be generated with a modulation index h of 0.5 and a BT product of 0.3. Note. The modulated signal is passed through a band-limited linear filter. The type of the filter used has a Gaussian impulse response in the time domain. Note. GMSK is a pre-filtered variant of the CPFSK which is used by UAT. It is popular in the commercial mass market and is used with different characteristics in GSM, DECT and TETRAPOL. Hence it offers significant advantage in cost reductions for the development and manufacture of avionics and ground-based equipment. Note. L-DACS2 will use GMSK as the baseline modulation with the same GMSK parameters as defined in [9]. Provisions are made to allow migration to higher order modulation schemes such as those used by GSM evolutions such as GPRS and EDGE, when the operational needs and traffic densities foreseen for Europe in the future warrant such evolution Modulation rate The modulating symbol rate shall be 1/T S = 1625/6 ksymb/s (i.e. approximately ksymb/s), which corresponds to 1625/6 kbit/s (i.e. approximately kbit/s), where T is the symbol period. Note. In GMSK, 1 symbol is equivalent to 1 bit. Note. One symbol period or 1 bit has a duration of T = 6/ s or approximately microseconds. Note. A logical 1 causes the carrier phase to increase by 90 o over a bit period and a logical 0 causes the carrier phase to decrease by 90 o. This phase change is produced by instantaneously switching the carrier frequency between two different values f1 and f2: f f 1 2 = f = f c c + R b R b / 4 / 4 (1) (2) where Rb is the modulation rate (1625/6 kb/s) and fc is the nominal carrier frequency Time/amplitude profile of L-DACS2 transmission The Reference Time shall be defined as the beginning of the first bit of the synchronisation sequence (the start of the Active part of the burst) appearing at the output port of the Edition: 1.0 Draft Page 23

24 equipment. Total burst duration Ramp up Sync sequence Start flag User data, FEC, CRC End flag Ramp down 8 bits µs 26 bits µs 8 bits µs 8 bits 8 bits µs µs Inactive transmitter Active part of the burst (bit transmission) Inactive transmitter Reference Time Figure 3: Time/amplitude profile of LDACS2 transmission Between the inactive state of the transmitter and the transmission of the first bit of the synchronisation sequence of a burst (the Reference Time), a ramp-up period of 8 symbol periods, equal to 8 bits or 48/ s (approximately microseconds) shall be respected. Note. The ramp up and ramp down period in currently available GMSK modulators are each 28 microseconds per [9]. In order to reduce transient spectral emissions during the ramp up and ramp down periods it may be necessary to increase the ramp up and ramp down times to limit the impact of the transients on the spectral footprint. It is anticipated that this may be one of the outcomes of the prototyping stage. For now, until prototype tests prove otherwise, it is recommended to use a ramp period at least as long as per [9], as this is known to be achievable in the GSM system. Prior to 8 bit periods before the Reference Time, the RF output power at the point at which the cable connects with the transmit antenna shall not exceed 80 dbm. Within 8 bit periods after the end of the Active part of the burst, the RF output power at the point at which the cable connects with the transmit antenna shall fall to a level not exceeding 80 dbm Ambiguity resolution and data transmission A synchronisation sequence consisting of 26 bits shall be transmitted following the ramp-up time. The synchronisation specifications are described in section 3.3. A start flag consisting of 8 bits shall be transmitted following the synchronisation sequence. The transmission of the first bit of user data shall start 42 bit intervals (approximately µs) after the nominal start of transmission. Note. Between the nominal start of transmission and the first bit of user data the 3-bit ramp up time, a 26-bit synchronisation sequence, and an 8-bit flag are transmitted. Note. The nominal start of transmission is the start of the ramp-up period of a burst. The start of the ramp-up period of a burst can in some sections of the frame coincide with the start of a slot; in other sections it can be delayed relative to the start of a slot due to a guard time being required to be maintained at the start of the slot; and in some sections of the frame the burst can be started ahead in time of the start of a slot in order that the burst will Edition: 1.0 Draft Page 24

25 coincide with the start of a slot upon reception, after propagation delay. A Cyclic Redundancy Code (CRC) field shall be transmitted after the user data field and used to provide detection of user data message errors. A Forward Error Correction (FEC) field shall be transmitted and used to provide correction of user data message errors. An end flag consisting of 8 bits shall be transmitted after the FEC, immediately prior to the ramp down period. Note. Some provisions for the CRC are made in the L-DACS\2 system in order to increase detection and correction capability. This CRC should be design at the transport layer in order to ask retransmission of the packet instead of the overall applicable message. This is out of the scope of the specification and should coordinate with cross layer QoS management even if provisions are made to implement the CRC. For the lower layer, the FEC gives sufficient protection to ensure the robustness of the link Receiver - transmitter turnaround time An L-DACS2 station shall be capable of beginning the transmission of the transmitter power stabilization sequence within 2 milliseconds after terminating the receiver function. An L-DACS2 station shall be capable of receiving and demodulating with nominal performance an incoming signal within 2 milliseconds after completing a transmission. Note. It is likely that a L-DACS2 ground station will be organised with a transmitter/receiver separated architecture Note. It is likely that an L-DACS2 airborne station will be organised with a transmitter/receiver separated architecture including two receivers to allow transparent cell handovers Frequency change during transmission The phase acceleration of the carrier from the start of the synchronization sequence to the data end flag shall be less than 300 Hz per second. 3.3 Air interface Radio frequencies The radio frequencies used shall be selected from the radio frequencies in the band MHz. The lowest assignable frequency shall be MHz (for spectrum compatibility) and the highest assignable frequency shall be 975 MHz. Note The 500 KHz guard band in the lowest part of MHz band aims to protect the telecommunication systems localized bellow 960 MHz against the L-DACS2 emission. The available spectrum shall be partitioned into a number of channels, each 200 KHz wide. Each of these bands shall be occupied by a GMSK modulated RF carrier supporting a number of TDMA time slots. The RF carriers may be aggregated in two or more paired combinations to accommodate different operational load in one cell. The separation between assignable frequencies that guarantees non interference between frequencies assigned to the same cell (channel spacing) shall be at least 600 khz. Note However, the separation between assignable frequencies can be reduced by imposing minimum clearance distance between the ground based antennas and the closest operating aircraft. The conditions and constraints under which this can be made possible are described in the Annex 5. Edition: 1.0 Draft Page 25

26 3.3.2 Channel bandwidth The L-DACS/2 channel shall have a nominal bandwidth of 200 khz for a single channel. For dense cells requiring high levels of capacity, the L-DACS2 channel may be combined in aggregations of two or more 200 khz channels. Note The process of channel aggregation may place practical constraints on the allowable distance between the ground station antenna and the nearest operating aircraft if no guard bands are used. The constraint exists in order for the ground station to be able to detect weak signals coming from aircraft far a field in the adjacent channel. Further information on these constraints is furnished in the Annex. Note The only constraint that exists whether adjacent channels are used or not, is the need to synchronise the two frequencies so as one is not trying to receive whilst the other transmits. In the absence of synchronisation, the transmissions from one channel may completely wipe out reception on the other channel. It has been found through simulation that a single 200 khz channel offers sufficient capacity to serve most operational scenarios. For scenarios requiring more than 132 kbit/s of pure user demand, two 200 khz channels will be required. Note This can be made possible by aggregating two (or more) channels as required, by sharing the loading of the cell across these channels. In practise, there are a number of ways to achieve this. One way is to perform segmentation of the cell using sectored antennas as is done in the GSM system to enhance the spectral efficiency of the system and simplify planning (as opposed to sharing a single omni-directional antenna). The following table summarises the channel bandwidth to support a single L-DACS2 cell under various loading conditions. These figures were obtained from the results of a campaign of simulations considering the various scenarios defined in the COCR. These scenarios are listed in the left column of the table. Scenario Services Number of aircraft TMA (large) Enroute (medium) Enroute (large - including AOC services 2 ) ATS and AOC ATS and AOC ATS and AOC User demand Channels required Total bandwidth kbps khz kbps khz kbps khz 4 Table 1 Total channel bandwidth for various cell sizes All the simulations have been carried out using 200 khz channels. The scenarios presenting the highest user demand were considered. 5 This includes the en route large scenario, the loading of which was split between two 200 khz channels with 102 aircraft assigned to each respective channel. The simulations results have thus shown a single 200 khz channel in the basic configuration to support all but the heaviest loaded scenario. In the latter case, two channels will be required. 2 This includes WXGRAPH service aircraft per channel. 4 Two 200 khz channels. 5 TMA large (52 aircraft), En route Medium (62 aircraft) and En route large (204 aircraft distributed across two 200 khz channels). Edition: 1.0 Draft Page 26

27 A summary of the latency and throughput performance offered in these scenarios is furnished in the following tables. These tables provide simulation statistics generate indicated evidenced in the caption of each table. The tables are divided into a number of columns headed by five latency requirements derived from [9]. These latencies are 95 th percentile figures and are comparable to the 95 th percentile figures achieved in the L-DACS2 simulations (figures in bold in the last row). Required Latency (s) (TT 95 ) Max latency (s) Average (s) th percentile (s) th percentile (s) Table 2 Latencies for TMA Large scenario (52 aircraft) Required Latency (s) (TT 95 ) Max latency (s) Average (s) th percentile (s) th percentile (s) Table 3 Latencies for En route Medium scenario (62 aircraft) Required Latency (s) (TT 95 ) Max latency (s) Average (s) th percentile (s) th percentile (s) Table 4 Latencies for En route large scenario (102 aircraft per channel) Scenario TMA LRG (52 aircraft) ENR MED (62 aircraft) ENR LRG (204 aircraft 6 ) Channel bandwidth Number of channels Achieved throughput 200 khz % 200 Khz % 200 khz % Table 5 Throughput achieved aircraft on each channel Edition: 1.0 Draft Page 27

28 3.3.3 Polarization L-DACS2 System Definition proposal: Deliverable D2 The design polarization of emissions shall be vertical Ground Frequency stability The radio frequency of L-DACS2 ground station equipment shall not vary more than ± % (1 parts per million) from the assigned frequency Aircraft Frequency stability The radio frequency of L-DACS2 aircraft equipment shall not vary more than ± % (5 parts per million) from the assigned frequency Spurious emissions Spurious emissions shall be kept at the lowest value which the state of the technique and the nature of the service permit. The spectrum mask is presented in the following table considering that the power output of L- DACS2 is higher than Power 43 dbm. For frequency offset below 1.8 MHz from the central frequency, the reference bandwidth is 30 khz, and for offset higher than 1.8 MHz from the central frequency, the reference bandwidth is 100 khz: Frequency offset from the carrier (khz) to to to to Attenuation threshold (dbc) On board Attenuation threshold (dbc) Ground Table 6 Spectrum mask (db) Note. This mask is achievable using a Surface Acoustic Wave (SAW) filter and a preselector filter before the amplificatory stage. A control of the output of the amplificatory stage shall guarantee the spectral quality of the emission. (Ongoing study evaluates the impact of this mask on the Eb/N0 needed to achieve a 10-3 BER, roughly estimated around 1 db). Note. L-DACS2 out-of-band emissions are expected to comply with ITU-R SM : the spurious domain consists of frequencies separated from the centre frequency of the emission by 250% of the necessary bandwidth of the emission. A reference bandwidth is a bandwidth in which spurious domain emission levels are specified. The following reference bandwidths are used: 100 khz between 30 MHz and 1 GHz from the carrier, 1 MHz above 1 GHz from the carrier. According to ITU-R SM , the maximum permitted spurious domain emission power in the relevant reference bandwidth is -70 dbc. The spectrum emission mask is described above is in line with the ITU regulation. The following table presents the ITU regulation on spurious emission: Edition: 1.0 Draft Page 28

29 Frequency offset from the central frequency Permitted spurious domain emission, dbm Reference bandwidth, khz f > f MHz or f < f MHz -70 dbc 100 Table 7 Spurious domain emissions ITU regulation Error phase specification For any 148-bits subsequence of the 511-bits pseudo-random sequence, defined in CCITT Recommendation O.153 fascicle IV.4, the phase error trajectory on the useful part of the burst (including tail bits), shall be measured by computing the difference between the phase of the transmitted waveform and the phase of the expected one. The RMS phase error (difference between the phase error trajectory and its linear regression on the active part of the time slot) shall not be greater than 5 with a maximum peak deviation during the useful part of the burst less than Broad band noise The Broad band noise shall be equal to -150 dbc/hz (relative to carrier) at frequency offset from the carrier > +/- 10 MHz. Note. This figure is sufficient for compatibility with GSM receiver. The minimum required distance between L-DACS2 Tx and a mobile GSM receiver is roughly estimated below 2 m Connection management The L-DACS2 system shall establish and maintain a reliable communications path between the aircraft and the ground system while allowing but not requiring manual intervention Transmission The physical layer shall encode the data received from the data link layer and transmit it over the RF channel. RF transmission shall only take place when it is permitted by the MAC Channel L-DACS2 channel shall be associated with one cell. L-DACS2 is a cellular system; frequency reuse pattern shall be implemented in order to optimise the coverage and the frequency band used. Note. First considerations toward the channel reuse and deployment is addressed in Annex Broadcast L-DACS2 should use specific channels for broadcast communications. Note. The L-DACS2 broadcast channels use a VDL Mode 4 modified MAC structure and frame structure (see Section 4.8). The requirements for Broadcast communication are not further discussed in this document Air-to-air point-to-point communication L-DACS/2 should use specific channels for air-to-air point-to-point communications. Note. The L-DACS/2 Air-to-air point-to-point channels use a VDL Mode 4 modified MAC structure and frame structure (see Section 4.9). The requirements for air to air point to point communication are not further discussed in this document. Edition: 1.0 Draft Page 29

30 System performance The system performance is designed to provide a required residual corrected Bit Error Rate (BER) of 10-7 on the basis of a Physical Bit Error Rate (BER) of Link Budget The system parameters retained for the link budget calculation are listed below: Edition: 1.0 Draft Page 30

31 UL DL TX Parameters Unit ENR ENR ENR TMA APT ENR ENR ENR TMA APT Governing Equation Notes L-DACS2 TX ouput Power dbm 55, , , , , , , , , ,9897 a Tx_Pout - (UL: 350W) - (DL: 50W) Maximum TX Antenna Gain dbi b TX_AntGain - (UL: DME antenna reference) - (DL: Antenna Gain = 3dB - polarization loss = 3dB) Tx Cable loss db 2,5 2,5 2,5 2,5 2, c TX_CableLoss - (UL: 50* *0.25) TX EiRP dbm 60, , , , , , , , , ,9897 d =a + b - c TX EiRP = TX_Pout + TX_AntGain - TX_CableLoss Propagation Parameters Transmit upper Frequency MHz e Tx-Rx Distance Nm f Path Loss db 143,62 139,18 133,16 129,64 117,60 143,62 139,18 133,16 129,64 g = 37,8 + 20log(f) + 117,60 20log(e) Free Space model (using nm unit) Miscellaneous Margins Interference Margin db h InterfMargin(TBD) Implementation Margin db i ImpMargin Safety Margin db j SafetyMargin(TBD) Banking Loss Margin db k Banking(TBD) RX Parameters Maximum RX Antenna Gain dbi l RX_AntGain - (UL: Antenna Gain = 3dB - polarization loss = 3dB) - (DL: DME antenna reference) Rx Cable loss (incl. Duplexer) db ,5 2,5 2,5 2,5 2,5 m RX_CableLoss - (UL: 50* *0.25) L-DACS2 RX receive signal dbm -82,68-78,24-72,22-68,70-56,66-91,63-87,19-81,17-77,65-65,61 n = d - g + l - m RxPower = TX_EiRP - PathLoss + Rx_AntGain - Rx_CableLoss Thermal Noise Density@290K dbm/hz o 10log(kT ) Bandwidth Hz p BW Thermal Noise Power dbm -120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99 q = o + 10log(p) 10log(k.T) +10log(BW ) Receiver Noise Figure db r Rx_NF Composite Noise Figure db ,5 9,5 9,5 9,5 9,5 z Composite noise including Rx noise and cable loss Total Rx Noise Power dbm -107,99-107,99-107,99-107,99-107,99-111,49-111,49-111,49-111,49-111,49 s = q + z + i + [h + j + k] N = Rx_NF + 10log(k.T.BW) BER=10-3 db t Eb/No L-DACS2 bit rate bps u R Required C/N db 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 v = t + 10log(u/p) C/N = Eb/No + 10log(R/BW) L-DACS2 Rx Sensitivity dbm -96,67-96,67-96,67-96,67-96,67-100,17-100,17-100,17-100,17-100,17 w = v + s Cmin = C/N + N L-DACS2 C/N available db 25,31 29,75 35,77 39,29 51,33 19,86 24,30 30,32 33,84 45,88 n-s L-DACS2 net db 8,00 12,43 18,45 14,97 27,02 2,54 6,98 13,00 9,52 21,56 n-s-v Considering ImpMargin and [other margins above] Figure 4 L-DACS2 Link Budget Edition: 1.0 Draft Page 31

32 The Tx output power is set at 350 W in the ground station and 50 W in the airborne equipment. This ground output power is achievable with a power amplifier and a power supply of 50 V. The ground cable loss has been fixed to 2.5 db using 50 meters of a 7/8 type cable (loss = 0.04 db/m) and 2 m of a RG214U (loss = 0.25 db/m) and connectors. The aircraft cable loss is set to 3 db in line with industrial input and L-DACS\1 assumptions. The L-DACS2 system is based on half duplex communication architecture (see Chapter 4). No diplexer either duplexer is needed to support the Rx and Tx function. An RF isolated switch shall be used to commute between Rx and Tx function. The Rx Tx turn around time is specified in section The antenna gain for the ground is 8 db, DME reference has been considered. The aircraft antenna gain is set to 0 db in line with the technology involved in UAT. The sensitivity considered in this link budget concerns the whole receiver block, receiver plus cable. The Banking loss is in line with L-DACS\1 assumptions. The safety margin has been set to 6 db. Note. This link budget doesn t take into account the antenna pattern for the different operational scenario presented above. Note. The impact of adjacent channel interference and co-channel interference has also been studied in order to confirm that the L-DACS2 system presents sufficient margin toward the impact of intra-system interference. This evaluation has been done in a statistical way and gives first ideas concerning the deployment of the system. The prototyping phase should confirm this analysis provided in Annex 4. Note. The link budget includes implementation losses. The prototype phase should show that extract net margin is sufficient to cope with fading losses and excess propagation losses. 3.4 Pulse blanking Current pulsed interference mitigation techniques fall under two categories: Time-domain approach, and Frequency domain approach. Pulse blanking is the time-domain method. It nulls out the portion where the amplitude of the signal exceeds a certain threshold level with respect to the noise. Pulse blanking has a number of advantages. It is simple to implement, it can be executed in real time without extra delay and can be activated only when the interference exists. However, when blanking the interference pulses, it also blanks out any other signals over that time slot. If the pulses are very dense in time, all received signals including DME/TACAN pulses and LDACS signals will be blanked, and data may be consequently lost due to unavailability of the signal. Furthermore, because of the Gaussian pulse tailing effect, pulse blanking cannot completely suppress the unwanted pulses and some residual interference will remain. Notch filtering is an alternative means of mitigating pulse interference in the frequency domain, where the pulsed signals e.g. DME and TACAN appear as narrow-band frequency tones. If the signal spectral density at certain frequencies is above the noise spectral density, these frequency components will be filtered out. Notch filtering can suppress DME/TACAN pulsed interference, including the central part of the Gaussian pulse and the tails. It also preserves the energy of the signal superposed with the interference pulses in the time domain. However, it not only filters interference, but also removes the signal energy at the Edition: 1.0 Draft Page 32

33 DME/TACAN frequencies. Even during the time period when there are no DME/TACAN pulses, the LDACS signal at these frequencies is still suppressed. If there are multiple DME/TACAN transponders nearby, the filter design may be excessively complicated due to multiple notches in the filter. Note. Hybrid blanking is an alternative technique developed for operation of GNSS in the L- Band that exploits the advantages of both pulse blanking and notch filtering. When an interference pulse is detected (in the time domain), it triggers the notch filtering of a brief slice (few microseconds) of data cantered at the estimated pulse position. Therefore, filtering is only implemented when DME/TACAN pulses exist. In doing so, it overcomes the disadvantage of regular notch filtering, which always filters out the corresponding frequency components of the signal even when there is no interference. For the portions of data that are covered by DME/TACAN pulses, hybrid blanking preserves most of the signal energy, and thus overcomes the disadvantage of time-domain pulse blanking. The filter design is simple, because there is only one notch in the filter. Although this technique has merits, it is more complex to implement than simple pulse blanking and the narrow band nature of the L- DACS2 (just 200 khz) makes this solution less viable than for wideband signals. Note. One of the key problems for a future communication component operating in the L- band ( MHz) is co-siting with other radio transceivers that operate in the same frequency band. Even if a frequency separation is implemented, providing some decoupling with other spurious emissions, the robustness of the new communication link will be drastically affected by the proximity of other pulse transmitters on the same aircraft Protection of L-DACS2 transmissions L-DACS/2 shall make use of Pulse Blanking Techniques, as necessary, to reduce the effect of strong interference sources (that is, the case on board aircraft due to very small system isolation) against L-DACS2 receiver damage. Note. Such a pulse blanking mechanism is defined in the UAT standards and has a common suppression bus interconnecting the avionics elements that could benefit from the information provided (pulse blanking signal whenever a transmitter is on). This primarily immunizes the receivers from co-site transmissions which would otherwise damage them or render inoperative for prolonged periods. Taking into account that the main sources of interference for the new communication system will be high powered, short pulsing transmitters (DME, UAT and SSR/Mode S), the duration of the jamming pulses will be equivalent to or shorter than the L-DACS2 symbol duration for which FEC data coding will be the appropriate answer to mitigate the impact of the interference on the frame error rate. Note. The typical suppression pulse for a UAT burst is 510 microseconds. The impact of the interference in his case will be therefore limited to a few bits in the frame for which FEC data coding will be the appropriate answer to mitigate the impact of the interference on the frame error rate. This means that the pulse blanking techniques are not required for limiting the interference on L-DACS2 but are useful to protect the L-DACS2 receiver from physical damages. The use of the suppression bus is one of a number of methods to implement pulse blanking. Therefore there is no hard requirement to define an interface to the suppression bus for the L-DACS2 receiving function Protection of adjacent systems transmissions L-DACS2 operates a power control scheme in order to limit interference on adjacent systems. Consequently, when operating in the vicinity of a ground station, the power emitted by the airborne transmitter is limited, and therefore unlikely to emit signals at damaging levels for other equipment on board the aircraft. Conversely, when operating close to the edge of the cell, the airborne radio radiates at maximum power, and the use of pulse blanking should be required to protect other L band receiving devices on board. Edition: 1.0 Draft Page 33

34 When irradiating at maximum power, the L-DACS2 transmitter outputs an estimated 47 dbm (50W). This is of sufficient strength to cause saturation of other L-band receivers, and potential damage to their front end circuitry. When operating at such power levels, it is therefore necessary that the L-DACS2 transmitter provides a pulse to activate the suppression bus during the transmission interval. Given the similarities in the power budget to that of UAT, it is considered feasible to use the same threshold cut-off of -20dBm for activation of the suppression pulses 7. The suppression interval required for L-DACS2 is thus derived on a similar basis to the UAT requirement for pulsed suppression. The L-DACS2 symbol period is approximately 3.69 microseconds (for 200 khz channels), which is of similar length to the pulses emitted by the DME and SSR systems. However, during the transmission of a burst, the L-DACS2 transmitter is active for a period of time during which the amplitude is stable and relatively constant. The amplitude stability means that over the course of the active transmission, the GMSK symbols in the payload part of the burst are transmitted within a constant power envelope. The length of the transmission itself is several orders larger than that of the pulsed DME and SSR systems and therefore unable to generate unsolicited SSR squitters, or false DME returns. The following table illustrates the impact of imposing a limitation to the transmission burst length on the latency performance offered for the various service levels in the reverse link. The results were obtained by simulating the heaviest available scenario (204 aircraft across two channels). A variety of burst lengths are considered ranging from a single slot up until eleven slots. Note The equivalent duty cycle is evaluated on a frame basis (one second), and provides a means of assessing transmitter active time as a percentage of the frame length. The overall channel loading due to user demand across a flight hour is considerably less (of order 0.3%). Burst length limitation (slots 8 ) Burst payload (bytes) Frame duty cycle (%) Latency requirements (s) Table 8 Effect of imposing burst length limits in the Reverse Link The above figures show that for most services, a burst length limitation of 1 slot will accommodate the majority of services within the latency requirement. However, the burst length is dimensioned by the largest transmitted messages of the FLIPINT and COTRAC services, as summarised in the table below. Service Size (Downlink) FLIPINT 2763 bytes 2.4 s COTRAC 1380 bytes 2.4 s Latency Requirement Table 9 FLIPINT and COTRAC messages sizes (downlink) 7 Per the UAT Manual, the -20 dbm requirement was derived from the maximum power allowable without SSR transponders generating unsolicited replies. These unsolicited replies occur when signal levels from the interferer signal envelope that lies within the SSR transponder receiver band is above the transponder receiver threshold. 8 Slots in the COS2 segment. Edition: 1.0 Draft Page 34

35 In these two cases, the minimum burst length possible is 9 slots and 5 slots respectively. For example, the 2763-byte FLIPINT message will be transmitted in 2 seconds (mod[2763/1549]) when using a 5 slot limit. However, since a fragment may occasionally be missed and thereby retransmitted, an 11 slot limitation may be necessary to accommodate the longest down linked messages within the 95th percentile latency requirement. Note. Data from simulations conducted on the basis of a large en - route scenario considering an L-DACS2 cell serving 204 aircraft were used to estimate the expected worst case duty cycle for the pulse blanking scheme. The simulation considered default L-DACS2 parameters. Given the rise time of 3 tail bits for the GMSK signal (equivalent to approximately 11 microseconds), it is considered feasible to use the same suppression tolerances of 5 microseconds required for UAT. The worst case maximum suppression interval allowed at any given time shall be 6.01 milliseconds including 5 microseconds before the start of the L- DACS2 transmission Interval, and 5 microseconds after the end of the same transmission. This suppression pulse shall be triggered when the L-DACS2 transmission exceeds -20dBm. Note. The validity of this threshold figure should be validated through trials. 3.5 Power control In order to reduce the level of interference for point-to-point, the L-DACS/2 transmitter shall have the capability of reducing the power of the transmitted signal using a power control function. This shall be done by using a small capacity in the signalling channels in the uplink frame section (afforded by the high capacity offered by the GMSK modulation option). Note. Power control may only be necessary for operation of L-DACS/2 in specific environments; however provisions are made in the signalling message to implement such a function. Airborne installation The effective radiated power shall be such as to provide an EIRP power output between 0.95 and V/m (200 mw to 50 W for the output power of the transmitter) on the basis of free space propagation, at ranges and altitudes appropriate to the conditions pertaining to the areas over which the aircraft is operating at 10 meters from the emission. This range of EIRP output is calculated to obtain the same level of Rx signal input from cell range of 10 Nm to 200 Nm. The airborne transmitter minimum power is 200 mw. The airborne transmitter maximum power is 50 W. Ground installation The effective radiated power shall be such as to provide an EIRP power output between 7.99 and V/m (2 W to 350 W for the output power of the transmitter) within the defined operational coverage of the facility, on the basis of free space propagation at 10 meters from the emission. This range of EIRP output is calculated to obtain the same level of Rx signal input from cell range of 10 Nm to 200 Nm. For the ground installation, the effective radiated power shall not change dynamically (see the note below). The ground station transmitter minimum power is 2 W. The ground station transmitter maximum power is 350 W. Note. The Ground station is supposed to reach all the mobiles in its operational range. Power control is not useful for the Uplink transmission. But if several stations are deployed to cover different airspaces (En-route, TMA, APT) the ground output power could be tuned to cope with the operational needs. Edition: 1.0 Draft Page 35

36 The power control management is described in section Received signals ground The ground base station shall perform a continuous measurement of the received signals from each aircraft and then return appropriate update power adjustment information to each aircraft ensuring the performance of link while optimising the power level at airborne side Received signals aircraft On reception of this information, the aircraft terminal shall use it to activate a power control function. Note. Advantage can be taken from algorithms developed and validated for GSM Single Antenna Interference Cancellation SAIC is a well known technology implemented in GSM to improve the interference mitigation and more precisely to improve the frequency reuse factor and the mobile load in each cell. A gain from 3 to 5 db for the co-channel interference ratio is foreseen in certain context using SAIC compliant mobile in GSM. Furthermore, this co-channel and adjacent channel interference mitigation technique has an efficiency improvement if the global network is synchronised. To cope with the L-band environment and to enforce the global capacity of L-DACS2 system, it is recommended that a SAIC feature is implemented in the L-DACS2 receiver equipment. Note. Different SAIC family algorithms exists Joint Demodulation (JD), also called Joint Detection and Blind Interference Cancellation (BIC). Both types of algorithm are well suited for GMSK modulation. It is observed from GSM simulation that the JD algorithm has better performance in interference mitigation than BIC algorithms for some scenarios but as a counter part JD algorithm complexities the receiver by 5 whereas the BIC algorithms induces a multiplication by 3 of the receiver complexity. Note. SAIC techniques have not been taken into account in L-DACS2 design. Anyway considering the potential benefit of SAIC, the prototyping stage should make provision to implement such techniques to evaluate or improve L-DACS2 spectral compatibility. 3.6 Interference immunity Transmitting function Adjacent channel emissions The amount of power from an L-DACS2 ground transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the first adjacent channel shall not exceed dbc. The amount of power from an L-DACS2 ground transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the second adjacent channel shall be less than -33 dbc. The amount of power from an L-DACS2 ground transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the third adjacent channel shall be less than -77 dbc and from thereon it shall monotonically decrease at the minimum rate of 5dB per octave to a maximum value of dbm. The amount of power from an L-DACS2 airborne transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the first adjacent channel shall not exceed dbc. Edition: 1.0 Draft Page 36

37 The amount of power from an L-DACA2 airborne transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the second adjacent channel shall be less than -33 dbc. The amount of power from an L-DACS2 airborne transmitter under all operating conditions when measured over a measurement bandwidth of 200 khz in the third adjacent channel shall be less than -77 dbc and from thereon it shall monotonically decrease at the minimum rate of 5dB per octave to a maximum value of -50 dbm Receiving function Specified error rate The specified error rate for L-DACS2 operation shall be the maximum uncorrected Bit Error Rate (BER) of 1 in 10 3 (0.1%) Reference sensitivity level The aircraft receiving function shall satisfy the specified error rate with desired signal strength of not more than dbm. The above specification shall be met by the aircraft receiving function with the two adjacent timeslots 20 db above the own timeslot. The ground receiving function shall satisfy the specified error rate with desired signal strength of not more than dbm. The above specification for BS shall be met when the two adjacent timeslots to the wanted are occupied with signals at 50 db above the power on the wanted timeslot. Note. The sensitivity figures specified in this chapter concern the whole receiver block, receiver plus Rx cable. Note. The above physical layer requirement is derived from [9] recommendations. Note. This considers static sensitivity in the absence of fading. Once fading conditions are included, although baseband dependant, typically 3 db extra signal-to-noise is required. Blocking tests carried out using GSM equipment have demonstrated good BER margin over the GSM specification of 2% even with a blocker at 2 db over the required level Adjacent band immunity performance and co-channel immunity performance The receiving function shall satisfy the specified error rate in section with a desired signal input level of 10 db above the reference sensitivity level, and for a random, continuous, GMSK-modulated interfering signal. The reference interference ratio for the ground and airborne receiving function shall be as defined in the following table. For co-channel interference for adjacent (200 khz) interference for adjacent (400 khz) interference for adjacent (600 khz) interference 9 db -9 db -41 db -49 db Figure 5 Adjacent band and Co-channel band reference interference ratio The interference ratio defined above lead to the evaluation of the adjacent frequency band and co-channel frequency band for frequency reuse and the use of multiple channels in one cell (see Annex 5). The minimum frequency guard band between 2 adjacent frequencies assigned to 2 adjacent cells shall be 400 khz. The minimum frequency guard band between 2 frequencies assigned in the same operational cell shall be 600 khz. Note. The actual interference ratio is defined as the interference ratio for which the BER of Edition: 1.0 Draft Page 37

38 10-3 is met Out-of-band immunity performance The airborne receiving function shall satisfy the specified error rate with a desired signal field strength of not more than dbm and with an undesired signal at least 59 db higher than the desired signal on any assignable channel 800 khz away from the assigned channel of the desired signal. The ground receiving function shall satisfy the specified error rate with a desired signal field strength of not more than dbm and with an undesired signal at least 59 db higher than the desired signal on any assignable channel 800 khz away from the assigned channel of the desired signal. Note. The range assumption used is 200 Nm Interference immunity performance The receiving function shall satisfy the specified error rate with a desired signal field strength of not more than -131 dbv/m, and with one or more out-of-band signals (except for VHF FM broadcast signals) having a total level at the receiver input of -33 dbm. 3.7 FEC mechanism The error coding scheme shall be based on concatenated code based upon convolutional coding (inner code) and RS coding (outer code). Note. The use of concatenated error coding is also considered to make the objective of a C/I of 9 db in co-channel interference attainable. Assuming a BER without coding at 10-3, the BER at the output of the RS coder shall be at least at Note. An interleaver schemes should be added. Two specific types of interleavers applicable to AMACS are block interleaving and diagonal interleaving. Further evaluation should be performed by running software simulations on a simplified transmission chain (no synchronization process, framing), with the appropriate transmission channel model, which could be the Ricean channel. Note. First discussion shown that the use of turbocode is not appropriated for short messages. Indeed the coding scheme is too heavy compared to the benefits. Another possibility is to code the whole uplink section as this section could be concatenated to form one long message. But this technique implies that one aircraft must receive the whole message before decoding. The decoding process is iterative and could have a non-negligible processing time. Therefore this idea was also rejected. A last idea to investigate could be to gather in the beginning of the uplink section all the sensitive signalisation information to be sent by the aircraft (framing message, Clear-To-Send (CTS) message and synchronisation parameters) and to protect it with a stronger coding scheme like turbocode. This method needs further investigation. Note. A discussion is provided in Annex 4 regarding coding and interleaving issues Inner code: convolutive punctured code The role of the convolutive code is to remove efficiently isolated errors. The well-known convolution code with a constraint length (LC) of (LC = 7, 171,133), already used with several puncturing scheme in DVB, will provide a range of performance regarding correction and coding rate. To be able to match the expected performances, the inner code coding rate shall be set to ¾ or 4/5. Decoding shall be soft-decision Viterbi decoder. Edition: 1.0 Draft Page 38

39 3.7.2 Interleaver L-DACS2 System Definition proposal: Deliverable D2 The interleaver depth shall be equal to one slot. This will ensure total independence of each communication. Note. The use of bloc interleaver or diagonal interleaver shall be assessed through proper simulations Outer code: Reed Solomon code Different RS codes have been considered, but because of: the relative length of a burst of error, the interleaver depth the length of a codeword RS(31,X,5) do not seem appropriate. RS(15,11,4) should provide better performance, because as the codeword is shorter, the errors will be spread over more codeword, leading to a relative smaller number of errors per codeword. A complete discussion is proposed in Annex 4. Edition: 1.0 Draft Page 39

40 CHAPTER 4 MAC sublayer 4.1 Introduction The MAC sublayer shall be the sublayer that acquires the data path and controls the movement of bits over this data path. The MAC sublayer shall acquire the shared communication path in order to provide the services defined in this chapter. Note. The functions performed by the MAC sub-layer should be transparent to higher functional layers. Note. For air/ground point-to-point communication L-DACS/2 has a high-integrity deterministic MAC sublayer, employing deterministic slot scheduling. Future extension to airair and broadcast functionality may employ self-organising TDMA principles in addition to deterministic slot scheduling. Note. The L-DACS2 frame length is designed for fast delivery of time-critical messages and has been set at 1 second. Simulations have shown the validity of this choice to achieve the highest performance requirements. Note. A QoS system is proposed to permit the use of channel resources according to the message transmit priority required. Specific channel slots are reserved for Request to Send (RTS) messages. This system allows the ground station, upon receipt of requests to send data corresponding to different priority categories, to prioritise the transmission of data with high priority Provision The MAC sub-layer shall provide: TDMA media access; time synchronization of the start of uplink and downlink slots in the channel; transmission processing; received transmission processing. Edition: 1.0 Draft Page 40

41 The MAC sub-layer shall receive from the L-DACS2 services sub-layer (LSS) a burst number for transmission, and the time of transmission. The MAC sub-layer shall provide the LSS with the received burst data, slot busy/idle status, slot occupancy status, signal level, and the status of the bursts sent for transmission MAC Layer for point-to-point and broadcast communication The MAC layer structure and usage for point to point communication shall be as specified in Sections 4.1 to 4.7. The MAC layer structure and usage for broadcast communication shall be as specified in Section 4.8. The MAC layer structure and usage for air to air communication shall be as specified in Section Framing L-DACS2 shall support a frame of duration M1 seconds, repeating every M1 seconds. A frame shall be equal in duration to M2 equally-spaced basic time slots. A frame shall consist of a number of frame sections. Each frame section shall be equal in duration to an integral number of basic slots, unless defined otherwise. A basic time slot shall have duration M5 = 1/M2 seconds. Note. A basic time slot may also be termed as a slot or a full slot. Note. A frame section may contain slots of shorter length than the basic time slot, provided that a frame section or a group of frame sections where so defined, has duration equal to an integral number of basic slots. The framing parameters shall have the default values as shown in Table 3. Parameter description Parameter name Value Frame duration (in seconds) M1 1 Number of basic slots per frame M2 150 Duration of 1 basic slot (in seconds) M5 1/150 Table 10: Parameter definition for framing parameters Note. The frame cycle is illustrated below. Note. The L-DACS2 frame length is designed for fast delivery of time-critical messages and has been set at 1 second. Simulations have shown the validity of this choice to achieve the highest performance. Frame n - 1 Frame n Frame n + 1 Frame cycle Frame cycle Frame cycle Figure 6 : L-DACS/2 frame Edition: 1.0 Draft Page 41

42 4.3 Synchronization L-DACS2 System Definition proposal: Deliverable D2 Note. For air-ground communications time synchronisation by mobiles based on measurement of the transmissions from ground station is sufficient. However the synchronisation specification for LDACS2 is designed to be compatible with future definition of an air-air service provided by the L-DACS2 system with system operation being possible air-to-air even if contact with the ground station is lost. Note. The L-DACS2 system shall be based on accurate time synchronisation between the ground station and aircrafts in range, in order to minimise the inefficient use of channel resources by the provision of a large guard time between each downlink slot. Note. The L-DACS2 synchronisation scheme operates in different modes to ensure the robustness of the synchronisation scheme. One of the synchronisation modes, the derived mode, is designed to be independent of external on-board time sources and is provided by the L-DACS2 system itself to avoid common mode failure between the L-DACS2 communication system and another avionics system. Note. In GSM, 6 bits are used for the timing correction and 26 bits for Timing Advance These 26 bits limit the coverage of 35 km in GSM. In the case of the GSM Timing Advance mechanism, the number of bits to code the Timing Advance and the correction is required to be in line with the maximum distance between the coverage edge of the cell and the ground station Specified time reference UTC time shall be specified as the time reference for station synchronization. Note. Each frame shall start at the beginning of a UTC second (see section 4.5) Primary time synchronization mode Under normal operating conditions, a mobile station shall maintain time synchronization such that the start of each successive group of M2 slots (a frame) is synchronized with the start of the specified time reference (see Section 4.3.1) second to within a 2-sigma value of 1 microsecond. Note. This is also defined as the primary time source Secondary synchronization mode When primary time is unavailable, a mobile station shall be capable of maintaining time synchronization such that the start of each successive group of M2 slots (a frame) is synchronized with the start of the specified time reference (see Section 4.3.1) second to within a 2-sigma value of 5 microseconds. Note. This is also defined as the secondary time source. Note. Secondary time is used only when the primary source has failed. A mobile station using secondary time shall however revert to using primary time whenever primary time is available Derived synchronization mode A mobile station shall be capable of deriving time synchronization from the framing message of the ground station with which it is in or wishes to make contact, such that the start of each successive group of M2 slots (a frame) is synchronized to within a 2-sigma value of 20 microseconds. Note. This is defined as the derived time source. Note. A mobile station will only be able to derive time from the framing message transmission of the ground station if the mobile station has knowledge of the distance Edition: 1.0 Draft Page 42

43 between the ground station and itself. To achieve this, the mobile station may have knowledge of its position from inputs from onboard position sources and may have a database containing ground station positions. Future message definitions for transmissions by mobile stations and ground stations may also include a requirement for the station positions to be transmitted on a regular basis (in the framing message), which may be recommended to aid timing functions and to be beneficial for other applications. Note. Derived time is used only when both the primary and secondary sources have failed Synchronisation of slots within a frame All slots within the frame shall be synchronised according to the frame structure provided by the ground station in the framing message (see section 4.5) Reversion A mobile station operating in the derived synchronisation mode shall revert to primary time whenever primary time is available. A mobile station operating in the derived synchronisation mode shall revert to secondary time whenever secondary time is available and primary time is not available. 4.4 Burst format Burst composition A burst shall be composed of: transmitter ramp-up period; synchronization sequence, start flag; user data (includes source and destination address if required); FEC, CRC code bits; end flag; transmitter ramp-down period; propagation guard time. A guard time of the specified duration shall be maintained between the end of the ramp-down period and the end of the slot or group of slots occupied by the burst. Note. All stations in the system maintain their own time reference, and maintain the frame structure and timing advertised by the ground station synchronised to this time reference. However all transmitted bursts suffer propagation delay in transit (unless the transmitter and receiver are co-located), such that all bursts arrive with some time delay after the slot start time defined by the framing structure. If a burst is transmitted at the start of a slot, the burst has to be shorter in time than the duration of a time-slot, to prevent the received burst overlapping a burst in the following slot (the following burst could be close to the start of the slot if the transmitter and receiver for that burst are closely located). The amount of time by which the transmitted burst is shorter than the duration of a slot is the guard time. The required duration of the guard time in this case is equal to the propagation delay that would occur for a mobile transmitting from the edge of the cell to the ground station (at the cell centre), which for this system is defined as 200 NM. The necessary inclusion of a guard time reduces the maximum length of bursts, and therefore restricts the net data rate that can be supported by the system. Combining multiple 1-slot messages into a single multi-slot burst helps to mitigate the loss of data caused by guard time, as only one guard time is required for a multi-slot burst. In selecting the slot duration to be used by the system, the required guard time is taken into account. Where a mobile station is making regular transmissions to a ground station, an alternative method is possible whereby: Edition: 1.0 Draft Page 43

44 a) The ground station measures the propagation delay of the first transmission by the mobile; b) The ground station indicates the propagation delay to the mobile in a dedicated message; c) The mobile uses the propagation delay information from the ground station to transmit its next burst to the ground station ahead of the scheduled slot start time by the amount of the expected propagation delay; d) The transmission from the mobile then arrives at the ground station at the slot start time, within a given uncertainty or error. e) The ground station continues to monitor changes in the burst arrival time from the mobile and continues to provide the correction information to the mobile. This method is referred to as adaptive burst alignment and the propagation delay information provided by the ground station to the mobile is referred to as the Timing Advance parameter (see Section 4.5.5). This method still also requires a guard time, but it allows the duration of the guard time to be reduced, thus permitting bursts to be longer than if the method is not used. The framing structure that is proposed for LDACS-2 used adaptive burst alignment in the parts of the frame where it is feasible to do so, and in the other parts the full guard time based on the size of the cell has to be used. An additional guard time is required between certain frame sections, and this is achieved with a guard time at the start of the uplink sections of the frame as well as at the end of the uplink sections. Details of the duration of the guard time required in different parts of the frame and of the dependence in this system of the guard time on the number of slots occupied by a burst are provided in Section For uplink bursts, a guard time of the specified duration shall additionally be maintained between the start of the slot or the first slot of a group of slots occupied by the burst and the start of the transmitter ramp-up period. Note. The burst structure for downlink bursts is illustrated in the figure below. The equivalent uplink burst is preceded by an additional guard period before the transmitter ramp-up period. Ramp up Sync sequence Start flag User data FEC, CRC End flag Ramp down Guard time Next slot 8 bits µs 26 bits µs 8 bits µs 8 bits 8 bits µs µs Active burst duration Total burst duration Total slot duration Figure 7: L-DACS/2 burst structure for downlink bursts Bursts occupying multiple slots If a ground station or a mobile station requires several concurrent slots for one transmission, Edition: 1.0 Draft Page 44

45 then the transmitted ramp-up time and synchronization interval shall only be present at the start of the initial slot, and the transmitter ramp-down time and propagation guard time shall only be present at the end of the last slot. Note. A single burst spanning two concurrent slots (and containing one message) is illustrated in the figure below for a downlink burst. The equivalent uplink burst is preceded by an additional guard period before the transmitter ramp-up period. Ramp up Sync sequence Start flag User data FEC, CRC End flag Ramp down Guard time Next slot 8 bits µs 26 bits µs 8 bits µs 8 bits 8 bits µs µs Active multi-slot burst duration Total multi-slot burst duration Total 2-slot duration Figure 8: L-DACS/2 merged slot structure for downlink bursts Where more than one message is contained in one burst, each pair of messages shall be separated by a one-octet flag. In this case, each message shall contain its own FEC error correction section of the message, within the flags bounding that message. Note. The number and position of the FEC code bits will be dependent on the size of the transmission. Note. A single burst containing two messages, with a flag separating the messages, is illustrated in the figure below for a downlink burst. The equivalent uplink burst is preceded by an additional guard period before the transmitter ramp-up period. Ramp up Sync sequence Start flag User data FEC, CRC Flag User data FEC, CRC End flag Ramp down Guard time Next slot 8 bits µs 26 bits µs 8 bits µs 8 bits µs 8 bits 8 bits µs µs Active multi-slot burst duration Total multi-slot burst duration Total 2-slot duration Figure 9: L-DACS/2 merged slot structure structure for downlink bursts Note. The number of user data octets that can be accommodated in a single burst spanning slots 1 to n is higher than the total number of user octets that can be accommodated if separate bursts are use to transmit the same data in slots 1 to n. Note. For a channel that is lightly loaded, a mobile station should aim to transmit data using Edition: 1.0 Draft Page 45

46 multiple (evenly spaced) single bursts to transmit data to the ground station, in order to minimise the interference effect on other aircraft systems that could be caused by the transmission of long bursts Ground station bursts The uplink sections are entirely organised by the ground station, and since the ground station has sole us of this portion of the channel, it transmits continuously in the UP1 and UP2 sections. Therefore the burst transmitted in a given uplink section could be (and most often is) several slots long. The least length an uplink burst could be is two slots, as illustrated in the figure below. The upper limit of the length is determined by the size of the UP1 or UP2 section. 9 As with multiple messages sent in downlink bursts, each message in the burst is accompanied by an FEC field. Figure 10: L-DACS/2 merged slot structure for uplink bursts 4.5 Framing structure Frame Structure A frame shall consist of two uplink sections, UP1 and UP2, and two downlink sections CoS1 and CoS2, and a login section LoG2. Note. The frame structure is illustrated in the figure below: 9 See further detail in Annex 1. Edition: 1.0 Draft Page 46

47 Start of UTC second 1 Frame UP1 CoS1 UP2 CoS2 Framing message Uplink section Downlink section LoG2 Login (Downlink) section Uplink section Figure 11: L-DACS2 frame structure (point-to-point) Downlink section In the LoG2 frame section, M3 LoG2 slots will exist in place of 1 slot. In the CoS1 frame section, M4 CoS1 slots will exist in place of 1 slot. A LoG2 slot shall have duration M6 = 1/(M2*M3) seconds. A CoS1 slot shall have duration M7 = 1/(M2*M4) seconds. The above framing parameters shall have the default values as shown in Table 3. Parameter description Parameter name Value Number of LoG2 slots per basic slot M3 2 Number of CoS1 slots per basic slot M4 6 Duration of a LoG2 slot (in seconds) M6 1/300 Duration of a CoS1 slot (in seconds) M7 1/900 Table 11: Parameter definition for framing parameters The UP1 and UP2 sections of a frame shall only be used by a ground station. The UP1 and UP2 sections of a frame shall be used by a ground station to send uplink messages to mobile stations in range, to send acknowledgement messages to mobile stations for data successfully received, to send Clear-To-Send (CTS) messages to mobile stations in response to Request-To-Send (RTS) messages received, and/or to send framing messages. The ground station shall concatenate all its messages into a single continuous burst, one for each UPx section (i.e. two per frame). Each of these bursts therefore spans a number of basic slots, with one burst to be transmitted in UP1 and one to be transmitted in UP2. The upper limit of the burst length is determined by the length of the UPx section. Each message in a ground station burst shall contain its own FEC section and shall be separated from a following message by one 8-bit flag. The CoS1, CoS2 and LoG2 sections of a frame shall only be used by mobile stations to transmit to a ground station. The length of each section of the frame shall be configured dynamically by each ground station. A ground station shall not be required to coordinate the lengths of the frame sections Edition: 1.0 Draft Page 47

48 (UP1, CoS1, LoG2, UP2, CoS2) with other ground stations. A mobile station shall listen to the framing message broadcast by a ground station in order to acquire knowledge of the framing lengths applicable to a particular ground station at a specific time. The UP1 section of the frame shall consist of M11 basic slots. The UP2 section of the frame shall consist of M12 basic slots. The CoS2 section of the frame shall consist of M14 basic slots. The CoS1 section of the frame shall consist of M13 CoS1 slots. The login section LoG2 of the frame shall consist of M15 LoG2 slots. Note. These parameters are defined in Table 6. The length of the CoS1 section plus the LoG2 section shall be equivalent to an integer number of basic slots. The CoS1 section of the frame shall be used by mobile stations to transmit in its guaranteed CoS1 slot (one slot per frame allocated by the ground station after the mobile s login insertion transmission, see Section 4.5.2). When a mobile station has data to transmit to the ground station, it shall transmit an RTS to the ground station to request slots to be reserved for data transmission in CoS2. If a mobile station has received data from the ground station it shall transmit an acknowledgement to the ground station in the next available CoS1 slot, which acknowledgement may be combined with an RTS. Each mobile allocated a CoS1 slot shall transmit a message in its CoS1 slot at least once per S1 frames. If the mobile does not have an RTS or acknowledgment to transmit to the ground station then it shall transmit a Keep-Alive message in its CoS1 slot at least once per S1 seconds in order to maintain transmissions in its CoS1 slot. Parameter description Maximum number of frames between transmissions by a mobile station in its allocated CoS1 slot Parameter name Default value Increment Range S to 30 Table 12: Parameter definition for S1 parameter The CoS2 section of the frame shall be used by mobile stations to transmit data to a ground station, either in slots that have been reserved for it by the ground station (reserved access), or in slots selected by the random access procedure defined in Section 5.6. Note. The number of slots assigned by random access and the number assigned by reserved access is variable. The number of slots assigned by reserved access should greatly exceed the number assigned by reserved access Login and response The LoG2 section of the frame shall be used by mobile stations to transmit a login message, using one LoG2 slot per frame, in one of the LoG2 slots made available by the ground station. A mobile station intending to transmit in the LoG2 section of the frame shall randomly select one of the available LoG2 slots using the slot selection algorithm defined in Section 5.7. The ground station shall allocate a CoS1 slot in the CoS1 section of the frame to each mobile station that has logged in to the ground station, for exclusive use by that mobile station while it remains within range of the ground station. The ground station shall provide to the mobile Edition: 1.0 Draft Page 48

49 station a unique local address both for itself and for the mobile station, to be used for all subsequent message exchanges between the ground station and mobile during the period that the mobile is logged in to the ground station. A login uplink response message shall be transmitted by the ground station in the UP1 section providing the slot allocation for each mobile station that has transmitted a login message in the login section LoG2 in the previous frame. Note. The login slots are used for initial contact when entering the cell and during the handover process. Cell exit exchanges are used during hand-overs (see section ). Login messages are used during hand-overs and during initial contact Frame section length alteration The length of each section of the frame shall be configured dynamically by the ground station according to the number of aircraft in range and the local demand for uplink and downlink capacity. The lengths of each section of the frame shall be broadcast in a framing message by the ground station, transmitted at the start of the UP1 section, at a rate of M16 times per minute. The ground station shall indicate a change to the frame section lengths, and provide the new section lengths, in the framing message, M17 frames in advance of the change to the frame section lengths. The ground station shall indicate in the framing message which frame number will be the one with the new section sizes by decrementing counter L1 in each framing message. Note. Parameters M16, M17 and L1 are defined in Table 13. Note. When the ground station intends to change the frame section sizes, mobiles with which it is in contact need to be notified multiple times in advance to ensure that all aircraft receive the frame change notification with a high probability. This implies notification several frames in advance Framing structure default parameter values and ranges The framing structure parameters shall have the default values and ranges as shown in Table 13. The default values shall only be used in the case of a ground station starting up having been previously switched off. Parameter description Number of slots in UP1 (basic slots) Number of slots in UP2 (basic slots) Number of slots in CoS1 (CoS1 slots) Number of slots in CoS2 (basic slots) Number of slots in LoG2 (LoG2 slots) Number of times per minute that framing message is sent Parameter name Default value (use only for starting from switched off state) Increment/ decrement Range M to 129 M to 129 M to 510 M to 145 M to 33 M to 60 Edition: 1.0 Draft Page 49

50 Number of frames in advance of a change to the frame lengths that the ground station is required to provide notification in the framing message Counter to be decremented prior to a change to frame section size M to 60 L to 31 M13/M4 + M15/M3 shall be an integer. Table 13: Parameter definition for framing parameters The number of slots in each frame shall equal M11 + M12 + M13/M4 + M14 + M15/M3. Note. The default size of the login section (LoG2) is based on the following calculation: for a typical ground station with a range of 200 NM and the maximum possible number of aircraft (204) distributed across the cell, there is calculated to be an average of 12 aircraft about to enter the cell in any one minute. If there are 4 LoG2 slots in each frame then over one minute there will be 240 slots available. The login sub-section of the frame shall always contain at least 4 LoG2 slots. This is to reduce the likelihood of different aircraft attempting to use the same slot if several handovers and/or initial logins are taking place simultaneously. Note. The number of slots in the CoS1 section should closely follow the number of loggedin aircraft. Note. Each aircraft is allocated one exclusive downlink slot to transmit an RTS in CoS1, allowing the request of slots in the CoS2 section, in order to guarantee fast access to the channel for messages requiring high QoS. This use of deterministic slot assignments is essential to achieve a high QoS performance Adaptive burst alignment A mobile station shall time the start of its login transmission with the slot start times indicated by the ground station for the LoG2 section of the frame. Note. The mobile maintains its own time, primary, secondary or tertiary. The ground station will indicate in its framing message a timeslot structure relative to the start of a UTC second. Note. The mobile does not use adaptive burst alignment (timing advance) for its login transmission in a LoG2 slot. Transmissions in the LoG2 section by the mobile will be initiated at the start of the LoG2 slot boundary, but will have suffered propagation delay by the time they reach the ground station. A ground station in receipt of a login message from a mobile station in one of the login slots shall measure the time difference, TD1 between the start of the slot as measured by the ground station and the time that the transmission from the mobile arrived. The ground station shall convert the time difference TD1 into a timing advance parameter T5 for that mobile according to the ranges specified in Table 15. The ground station shall include the value of the timing advance parameter T5 for each mobile in a timing advance message that is transmitted to the mobile in response to its login message. Parameter description Parameter name Default value Increment Range Timing advance parameter T to 96 Table 14: Parameter definition for timing advance parameter T5 Edition: 1.0 Draft Page 50

51 For each mobile, once the ground station has transmitted the first timing advance message to the mobile, the ground station shall monitor the timing of all the transmissions from the mobile in CoS1 and CoS2 for the purpose of assessing if the T5 parameter provided to the mobile needs to be changed. Note. If a mobile does not have any data or acknowledgment to transmit to transmit to the ground station, a Keep-Alive synchronisation message must be sent to the ground station at least once per S1 frames (see Section 4.5.1). A ground station in receipt of a burst from a mobile station in CoS1 or CoS2 shall measure the time difference, TD2 between the start of the slot for the ground station s own framing and the time that the transmission from the mobile arrived. The ground station shall convert the time difference TD2 into a timing advance parameter T5 for that mobile according to the ranges specified in Table 15. Using the measurement of TD2, the ground station shall, at least every S1 seconds, for each mobile, indicate whether the mobile must increase or decrease its timing advance parameter T5, and by what amount, or whether it must leave its T5 parameter unchanged. Note. In LDACS2, a one-way measure of time difference due to propagation delay is used, as opposed to a two-way measure, since the LDACS2 mobile stations are required to maintain their own time, synchronised to UTC. In systems where mobiles derive time from the ground station, a mobile s transmissions are delayed by the time they reach the ground station by an amount equivalent to twice the propagation distance (the mobile would measure the start time of a frame from the transmissions by the ground station the mobile s frame structure is delayed by an amount due to the one-way delay from the ground station). Note. The timing advance parameter has been specified assuming that aircraft speeds for aircraft using this communication system are no higher than 600 knots (Mach 1 at 25,000 ft under standard conditions). With this assumption, a message needs to be transmitted at least once per 30 seconds from each mobile to the ground station to allow the ground station to assess, based on variations in the timing of the mobile station s burst start time with respect to the timing expected (from the current T5 value allocated to the mobile), whether a modified T5 value needs to be provided to the mobile. A maximum 30 second update period for T5, accommodated by the maximum setting for S1 of one transmission in CoS1 per 30 frames, provides an appropriate balance between changes in the synchronisation of the mobile bursts with the ground station slot structure and the load caused by the regular update of T5 in the ground station to mobile communication exchanges. A default setting for S1 of one transmission in CoS1 per 10 frames provides increased assurance that aircraft are provided with the correct T5 value corresponding with their distance from the ground station, considering that some transmissions may be lost in either direction between ground station and mobile. Note. A mobile station may be able to verify the correctness of the T5 value provided to it by the ground station if it has knowledge of the distance between itself and the ground station. The mobile station may be able to use this information to filter out isolated T5 values that are obviously incorrect. However it should not use this information to disregard more than one T5 value at a time, as the assumption should be made that the ground station always has an accurate measure of the times of arrival of mobile transmissions with respect to the slot start times measured by the ground station. Range of TD1/TD2 Corresponding range between ground station and mobile Allocated T5 value Time by which mobile advances transmission 0 < Tdiff <= ms 0 < R <= NM ms < Tdiff <= ms NM < R <= 6.25 NM ms ms < Tdiff <= 6.25 NM < R <= ms Edition: 1.0 Draft Page 51

52 ms NM ms < Tdiff <= ms NM < R <= 12.5 NM ms continued until continued until continued until continued until ms < Tdiff <= ms ms < Tdiff <= ms NM <= R ms NM <= ms ms < Tdiff 300 NM < R ms Table 15: Allocation of T5 parameter to transmission delay Burst transmission in relation to slot and frame structure UP1 and UP2 sections For bursts transmitted by a ground station in UP1 or UP2, the start of the transmitter ramp-up period of a burst shall begin after a start guard period of 408 bits (approximately ms) from the start of the UP1 or UP2 frame section. Note. The guard period at the start of the UP1 and UP2 frame sections is needed to provide a 1.5 ms transmit to receive turnaround time for the mobiles and a 1.5 ms receive to transmit turnaround time for the ground station. For bursts transmitted by a ground station in UP1 or UP2, a guard period of length indicated in Table 7, in which no transmission is made by the ground station, shall be reserved between the end of the ramp down period and the end of the slot. Number of slots occupied by burst in UP1 or UP2 (n = integer greater than zero) Length of concurrent slot group in bits Guard period to be provided at the end of a burst in UP1 or UP2 in bits Approximate duration of guard period to be provided at the end of a burst in UP1 or UP2 9n 8 (9n 8)*16250/ / ms 9n 7 (9n 7)*16250/ / ms 9n 6 (9n 6)*16250/ / ms 9n 5 (9n 5)*16250/ / ms 9n 4 (9n 4)*16250/ / ms 9n 3 (9n 3)*16250/ / ms 9n 2 (9n 2)*16250/ / ms 9n 1 (9n 1)*16250/ / ms 9n (9n)*16250/ ms Table 16: End guard period for UP1 and UP2 burst as a function of the number of slots occupied by the burst Note. The guard time required is based upon the following considerations: a) Following transmission of a burst by a ground station in UP1 or UP2, upon reception at the mobile station the burst will be delayed with respect to the start of the slot by the time taken for the burst to travel from the ground station to the mobile station. The guard period therefore needs to provide protection from burst overlap for ground station to mobile station distances that are likely with a 200 NM radius cell. To provide 200 NM protection, the guard period must exceed milliseconds. b) At the end of UP1 or UP2, the ground station needs a transmit to receive turnaround time of 1.5 ms, and mobiles transmitting at the start of CoS1 or CoS2 (which follow UP1 and UP2 Edition: 1.0 Draft Page 52

53 respectively) need a receive to transmit turnaround time of 1.5 ms, which leads to a need to extend the guard time by 1.5 ms. c) Within the 30 s timing advance parameter update period, a mobile transmitting in the first slot of CoS1 or CoS2 could travel closer to the ground station, causing a burst to overlap either the UP1 or UP2 sections by up to ms, and could have a clock that is ahead of that of the ground station by up to ms, leading to a need to extend the guard time by a further ms. d) The guard time at the end of an UP1 or UP2 slot thus needs to be in excess of = ms e) The guard period is then adjusted above this value depending on the number of slots occupied by the ground station burst, in order to keep the useable part of any burst equal in length to a whole number of bits LoG2 insertion section For login bursts transmitted by a mobile station in the LoG2 section of the frame, the start of the transmitter ramp-up period of a burst shall be aligned with the start of a time slot at the time of transmission. For login bursts transmitted by a mobile station in LoG2, a guard period of /9 bits (approximately bits or ms) in which no transmission is made by the mobile station shall be reserved between the end of the ramp down period and the end of the slot. Note. The guard time required is based upon the following considerations: a) Following transmission of a burst by a mobile station in LoG2, upon reception at the ground station the burst will be delayed with respect to the start of the slot by the time taken for the burst to travel from the mobile station to the ground station. Therefore in the LoG2 sections of the frame, a guard time sufficient to provide protection from burst overlap for mobile ranges at least out to the cell radius of 200 NM needs to be provided, for which a guard time above milliseconds is required. b) Two mobiles transmitting in adjacent slots in the LoG2 section could each have a clock running in error by up to 0.02 ms, and thus 0.04 ms needs to be added to the guard period. c) The above results in a requirement for a guard time of at least ms. d) The guard period is then adjusted above this value, in order to keep the useable part of any burst equal in length to a whole number of bits CoS1 and CoS2 sections For bursts transmitted by a mobile station in CoS2 and in CoS1, the start of the transmitter ramp-up period of a burst shall begin in advance of the scheduled slot start time at the ground station by an amount indicated by the ground station to the mobile in the Timing Advance Parameter T5 (see Section 4.5.5). Note. In the CoS1 and CoS2 frame sections, the mobiles transmit using a feature called Timing Advance (used in GSM), in order to obtain a synchronised frame at the ground station while at the same time minimising the required length of guard time to be allowed for at the end of a burst. Timing Advance cannot be used by the mobiles in the LoG2 section because prior to logging in to the ground station the mobiles do not have knowledge of the mobile to ground station propagation delay CoS1 section For bursts transmitted by a mobile station in CoS1, a guard period of /27 bits (approximately bits or ms) in which no transmission is made by the mobile station shall be reserved between the end of the ramp down period and the end of the slot. Note. The guard time required is based upon an estimated maximum burst overlap at the ground station for two mobile stations, A and B, transmitting in adjacent slots using timing Edition: 1.0 Draft Page 53

54 advance, plus maximum errors in clock synchronisation for these two mobiles, as follows: a) Mobile A transmits in the first of the two slots; mobile B transmits in the following slot; b) Maximum additional propagation delay of mobile A s burst due to mobile A having the longest propagation delay compatible with the allocated T5 value: milliseconds; c) Maximum additional propagation delay of mobile A s burst due to mobile A having travelled away from the ground station at 600 knots in the 30 seconds that have elapsed since the T5 value was allocated to the mobile (600 knots is Mach 1 at 25,000 ft under standard conditions): milliseconds; d) Maximum additional delay of mobile A s burst due to the clock of Mobile A being slow and at the limit of the allowed synchronisation tolerance: milliseconds; e) Maximum additional delay for burst of Mobile A: milliseconds; f) Maximum reduction in propagation delay of mobile B s burst due to mobile B having the shortest propagation delay compatible with the allocated T5 value: 0.0 milliseconds; e) Maximum reduction in propagation delay of mobile B s burst due to mobile B having travelled towards the ground station at 600 knots in the 30 seconds that have elapsed since the T5 value was allocated to the mobile (600 knots is Mach 1 at 25,000 ft under standard conditions): milliseconds; f) Maximum reduction in delay of mobile B s burst due to clock of Mobile B being fast and at the limit of the allowed synchronisation tolerance: milliseconds; g) Maximum reduction in delay for burst of Mobile B: milliseconds; h) The maximum burst overlap for two mobiles using adjacent slots if there was no guard time would be: = milliseconds. The guard period is then adjusted above this value, in order to keep the useable part of any burst equal in length to a whole number of bits CoS2 section A mobile station intending to transmit in the CoS2 section of the frame shall transmit bursts occupying between 1 and M21 slots, where the value of M21 is defined in the Table 17. The value of M21 shall be broadcast by the ground station in its framing message. Parameter description Maximum length of burst to be transmitted by a mobile station in CoS2 (in slots) Parameter name Default value Increment Range M to 10 Table 17: Parameter definition for maximum burst length in slots in CoS2 For bursts transmitted by a mobile station in CoS2, a guard period of length indicated in Table 18, in which no transmission is made by the mobile station, shall be reserved between the end of the ramp down period and the end of the slot. Edition: 1.0 Draft Page 54

55 Number of slots occupied by burst in CoS2 (n = integer greater than zero) L-DACS2 System Definition proposal: Deliverable D2 Length of concurrent slot group in bits Guard period to be provided at the end of a burst in CoS2 in bits Approximate duration of guard period to be provided at the end of a burst in CoS / / ms 2 2*16250/ / ms 3 3*16250/ / ms 4 4*16250/ / ms 5 5*16250/ / ms 6 6*16250/ / ms 7 7*16250/ / ms 8 8*16250/ / ms 9 9*16250/ ms 10 10*16250/ / ms Table 18: Guard period for CoS2 slot as a function of slot length Note. The guard time required is based upon an estimated maximum burst overlap at the ground station for two mobile stations, A and B, transmitting in adjacent slots using timing advance, plus maximum errors in clock synchronisation for these two mobiles. The derivation in Section showed that the maximum burst overlap for two such mobiles using adjacent slots if there was no guard time would be: = milliseconds. The guard period is then adjusted above this value, in order to keep the useable part of any burst equal in length to a whole number of bits Summary of burst guard times Note. Table 10 summarises the guard times applicable to each section of the frame, for guard periods to be maintained after the end of each burst (after the 3 bit ramp down period). Section of frame Section of frame following Tx to Rx / Rx to Tx time needed at end of section Guard time to protect to 200 NM Burst overlap from timing advance and/or mobile clock error Minimum duration guard period must satisfy Allocated actual guard period in bits UP1 CoS1 1.5 ms ms ms ms See Table 5 in Section CoS1 LoG2 not required (see Note) LoG2 UP2 not required (see Note) not applicable ms ms /18 bits (See Section ) ms 0.04 ms ms /9 bits (See Section ) UP2 CoS2 1.5 ms ms ms ms See Table 5 in Section CoS2 UP1 not required (see Note) not applicable ms ms See Section Approximate actual guard period duration See Table 5 in Section ms ms See Table 5 in Section Note. In the case of the Log2 and CoS2 frame sections, there is no requirement for an additional guard time at the end of these sections, since the Tx to Rx / Rx to Tx turnaround time is catered for by a ms guard time at the start of the UP1 and UP2 frame sections (see following table). In the case of the CoS1 section, this additional guard time is not required because the CoS1 section is Edition: 1.0 Draft Page 55

56 followed by a further downlink section. L-DACS2 System Definition proposal: Deliverable D2 Table 19: Summary of burst guard periods per frame section Note. Table 20 summarises the guard times applicable to each section of the frame, for additional guard periods to be maintained at the start of a frame section. Section of frame Section of frame following Tx to Rx / Rx to Tx time needed at start of frame section Allocated actual guard period at start of frame section in bits Approximate actual guard period duration at start of frame section UP1 CoS1 1.5 ms 408 bits ms CoS1 LoG2 not required Tx to Rx / Rx to Tx is catered for by guard time at end of UP1 - - LoG2 UP2 not required - - UP2 CoS2 1.5 ms 408 bits ms CoS2 UP1 not required Tx to Rx / Rx to Tx is catered for by guard time at end of UP1 - - Table 20: Summary of burst guard periods at the start of frame sections 4.6 Quality of Service (QoS) management Efficiency Note. The efficient handling of QoS is based on the TDMA structured MAC layer and provides transmission based on data priority L-DACS2 QoS management Parameters The L-DACS2 system shall permit handling of QoS based on the Priority parameter Q22. Parameter description Parameter name Default value Range Priority Q to 3 Table 21: Definition of parameter Priority Q22 Message category ATN Priority Network/systems management 14 3 Distress communications 13 3 Urgent communications 12 3 High priority flight safety messages 11 3 Normal priority flight safety messages 10 2 Meteorological communications 9 2 Flight regularity communications 8 2 Aeronautical information service messages 7 2 Network/systems administration 6 2 Aeronautical administrative messages 5 1 Q22 value Edition: 1.0 Draft Page 56

57 Unassigned 4 1 Urgent priority administrative and UN charter communications 3 1 High priority administrative and state/government communications 2 1 Normal priority administrative 1 0 Low priority administrative 0 0 Table 22: Mapping between message category, ATN priority and Q22 priority classification Note. This mapping is an initial proposition. It could evolve with the definition of the QoS management with end to end perspective (e.g. revision of the ATN priority table). The L- DACS2 resource allocation mechanism is design to meet this future QoS management requirement (deterministic/best effort resource allocation) Priority information Note. The priority flag Q22 shall be used to distinguish the relative importance of the exchanged data within a given QoS (best effort or guaranteed) with respect to gaining access to communications resources and to maintaining the requested QoS. Note. The priority of different message categories has been specified by ICAO in terms of the ATN priority (a subset of those categories belongs to the guaranteed QoS). The definition of the Q22 priority parameter is based on the ATN priority. On sending data to be queued for transmission the Priority Q22 shall be set to the appropriate Q22 value for the message content as indicated in Table 7. A station shall prioritise transmission of data with higher Priority Q22 value over data with lower Priority Q22 value. When L-DACS2 has multiple messages queued to send with different Q22 priorities, then it shall take account of the Q22 priority in deciding which messages to send first. When a capacity request has been done for a message and a new message arrives in the sending queue with a higher Q22 priority it shall be transmitted first (if possible according to reservation size) End-to-end QoS management Note. End-to-end communication will involve heterogeneous networks, including mainly an air-ground link (typically L-DACS2 radio link or another equivalent radio link) and a ground transport network. Management of QoS on the L-DACS2 link has been addressed in Section In order to be able to provide end-to-end QoS management between the airborne system and the ground controller system, two alternatives are envisaged: Implementation of QoS management mechanisms on the ground network infrastructure. Solutions based on IP based infrastructure, using IntServ or DiffServ model, are envisaged. Implementation of QoS management mechanisms at the transport level. This transport protocol shall be designed to be used over a network layer that provides best-effort service differentiation (called EDS Equivalent Differentiated Services). This solution has the advantage of providing this information directly to end users in order to decide whether the communication infrastructure is capable of providing the expected QoS. These options should be investigated within the SESAR development phase within the project Processing Bursts received from the MAC sub-layer shall be forwarded to the physical layer, along with the time for transmission. Edition: 1.0 Draft Page 57

58 4.8 Power Control management The L-DACS2 airborne power control mechanism shall be supported by a power control parameter send by the ground station to the mobile logged in its service area. This parameter encoded on 5 bits shall code the range of possible airborne power control output as described in section 3.5. From section 3.5, the airborne output power could be adapted from 23 dbm to 47 dbm. The power control parameter encodes the 25 possible states of the output power using a 1 db step among the 32 available states allowed with 5 bits. If the desirable output power is 23 dbm, the ground station set the power control field to If the desirable output power is 47 dbm the ground station set the power control field to corresponds to not monitor. The ground station shall send a message to update the airborne output power as soon as the ground station detect that an evolution in the ground received power sufficiently significant to change the state of the transmitted power. If the received power is bellow -90 dbm, the ground shall increase the power control parameter from one state. If the receiver power is above -89 dbm during 5 seconds, the ground station shall decrease the power control parameter from one state. The ground station shall use the power control field available in the uplink message if one message will be sent to the mobile or use the power control field available in the uplink power control message (see Annex 2). 4.9 MAC layer for Broadcast Service Frame The broadcast service framing shall be as defined in Section Synchronisation The broadcast service synchronisation shall be as defined in Section Burst format The broadcast service burst format shall be as defined in Section Frame structure The broadcast service frame structure shall start with a ground-quarantine section of length B1 slots for fixed access by ground stations. The remaining slots in the frame shall be available to all stations (mobile and ground) by reserved and random access. The fixed access, reserved access, and random access protocols shall be based on modified VDL Mode 4 fixed access, reserved access, and random access protocols. Note. Most VDL Mode 4 broadcast protocols will be used, but no point-to-point transmissions will be permitted on the broadcast channel. Therefore some modifications will be required. Note. Further definition of the broadcast service will take place at a later stage MAC layer for Air-to-Air Point-to-Point Service Frame The air-to-air point-to-point service framing shall be as defined in Section 4.2. Edition: 1.0 Draft Page 58

59 Synchronisation L-DACS2 System Definition proposal: Deliverable D2 The air-to-air point-to-point service synchronisation shall be as defined in Section Burst format The air-to-air point-to-point service burst format shall be as defined in Section Frame structure The air-to-air point-to-point service frame structure shall start with a ground-quarantine section of length B1 slots for fixed access by ground stations. The remaining slots in the frame shall be available to all stations (mobile and ground) by reserved and random access. The fixed access, reserved access, and random access protocols shall be based on modified VDL Mode 4 fixed access, reserved access, and random access protocols. Note. Most VDL Mode 4 air-to-air point-to-point protocols will be used, but some modifications will be required. Note. Further definition of the air-to-air point-to-point service will take place at a later stage. Edition: 1.0 Draft Page 59

60 CHAPTER 5 Data link sublayer (DLS) 5.1 Introduction The data link service (DLS) sublayer shall be the sublayer that manages the transmit queue, creates and destroys Data Link Entities (DLEs) for connection-orientated communications, provides facilities for the LME to manage the DLS, and provides facilities for connectionless communications. The data link sublayer (DLS) shall support communications on a shared communications channel as described in this chapter. The DLS shall provide the following services transmission of user data, indication that user data has been sent, reception of user data, indication that the DLS link has been established, and indication that the DLS link has been broken. The DLS shall rely on the MAC layer to ensure that messages corrupted during transmission are detected and discarded. 5.2 Transmission procedure The user data shall be split up into separate segments if the message size exceeds the maximum that can be accommodated into the first available slot or slots. The first segment will be transmitted in the first available slot or slots. Note. The maximum message size for one slot in CoS2 corresponds approximately to 158 octets of user application data (see Section 3.4). Note. For each additional consecutive slot that is available over and above a single slot, approximately an additional 173 octets of user data can be transmitted in the combined slot window. This increment in user data greater than the number of octets that can fit in a single Edition: 1.0 Draft Page 60

61 slot as a result of savings from not having to include additional flags, addresses or guard time. Each segment shall carry a sequencing marker to indicate both the total number of segments and each segment s number in the sequence. A message identifier field shall be included in each message in addition to the message type field. Note. The message identifier field ensures that stations can be certain as to which of their transmissions have been acknowledged. Note. For example if a long data message is transmitted in several segments and one segment is not correctly received it would be inefficient to have to retransmit the whole message. The message identifier shall be a rolling sequence number, with values in the range 1 to 63. Note. It is not necessary for every message to have a unique ID, merely for the messages from a station to be distinguishable within a period of time. The receiving station shall include the message identifier of the received message in its acknowledgement, in order to indicate which message is being acknowledged Uplink transmission procedures Note. The UP1 and UP2 sections of the frame are reserved for uplink transmissions. These blocks are to be used by the ground station for normal data transmissions to aircraft, acknowledgements to downlink messages, CTS messages, framing messages and login and cell exit exchanges. The slots in the uplink sections shall be concatenated and shall not require separate ramp-up and ramp-down times nor guard-times in between messages. If, at the start of the UP1 section, the ground station has a data message or multiple data messages to send, then, provided there is space available in the UP1 section, it shall concatenate the data message or messages with any framing message or login response message or combined ACK/CTS message and transmit the concatenated set of messages in one single burst in UP1, using an 8-bit flag to separate each message. If, at the start of the UP2 section, the ground station has a data message to send or multiple data messages, then, provided there is space available in the UP2 section, it shall concatenate the data message or messages with any combined ACK/CTS message and transmit the concatenated set of messages in one single burst in UP2, using an 8-bit flag to separate each message Downlink transmission procedures If, at the time the aircraft has the opportunity to transmit in its allocated CoS1 slot, the mobile station has data in its queue for downlink, the aircraft shall transmit a Request To Send (RTS) reservation request for CoS2 slots in its CoS1 slot indicating the number of slots required, and the priority of the messages to be transmitted. On reception of the RTS, a Clear To Send (CTS) shall be transmitted by the ground station in the UP2 section. If slots are available in the CoS2 section, the CTS shall acknowledge the request for time slots and shall indicate which slots have been allocated in CoS2. The ground station shall reserve slots in CoS2 starting with the earliest time slots in CoS2 being used first. If CoS2 slots have been reserved, the mobile station shall transmit the data in its allocated CoS2 slots. If no slots are available in the CoS2 section, the CTS shall acknowledge the request for time slots and shall indicate that no slots are available. Edition: 1.0 Draft Page 61

62 If some, but insufficient slots, are available in the CoS2 section, the CTS shall indicate those slots that have been reserved and acknowledge that the request remains in place for further time slots to complete the data transfer. If no or insufficient slots are available but the CTS has acknowledged that the request for slots remains in place then the mobile station shall issue a further RTS in its CoS1 slot in the next frame, adding in any additional requests that it may now have as a result of a demand for transmission of further messages. If no CTS is received by the aircraft, or if the ground station replies with a NACK, then the aircraft shall attempt to transmit the data by random access in accessible slots in CoS2. If a CoS2 data transmission from the aircraft is not acknowledged by the ground station in the next frame, or if the ground station replies with a NACK, then the aircraft shall issue a further RTS to the ground station using the CoS1 slot in the next frame. If a mobile station receives queued data to send after the start of its allocated CoS1 slot and before the end of the CoS2 section, and it knows that there are available slots in CoS2,, it shall attempt to transmit the data by selecting slots for random access transmission using the random access transmission algorithm specified in Section 5.6. In order to prevent conflicting transmissions, all aircraft shall listen to all CTS transmissions to record in their reservation tables which slots have been reserved (in all sections of the frame). 5.3 Reception procedure On reception of user data blocks from the MAC layer, the DLS shall determine (from the sequencing numbers) how many blocks to expect. The re-assembly of the blocks shall be done by using the sequencing numbers. 5.4 Segmentation The DLS shall handle the segmentation of user data queued for transmission by higher layers into appropriate blocks for the MAC layer. The DLS shall handle the de-segmentation (re-assembly) of received blocks from the MAC layer into a single user data packet for the upper layer. 5.5 Reserved access protocol specification Every station (air and ground) shall keep a table of all known stations. For each station, the table shall include: the type of the station; the station s local address; a copy of the last type of transmission; the time of the last transmission. Note. The details and processing of this table are dependent upon the implementation adopted by the radio manufacturers. A station shall maintain a table of all reservations in the current frame. For each reserved slot, the reservation table entry shall consist of the local address and (when available) 27-bit address of the intended transmitter, the local address and (when available) 27-bit address of the destination (if any) and the type of reservation made. Edition: 1.0 Draft Page 62

63 An aircraft s reservation table shall be updated during each frame after the receipt of the ground station s uplink messages. An aircraft shall also update its reservation table on receipt of reservation messages from other aircraft or from the ground station. When a station has a data to transmit for which it has a reservation, it shall transmit the scheduled data in the reserved slots. Note. The reserved slots which are used shall depend on the identity of the station and the amount of data to be sent. Note. The details and processing of this table are dependent upon the implementation. 5.6 Random access protocol specification for transmission in CoS2 When there is sufficient demand for slots in CoS2, the ground station shall, where possible, satisfy the demand for slots by allocating reserved slots to mobiles using the reserved access protocol. Note. The ground station may allocate all the slots available in COS2 using the reserved access protocol. Use of the reserved access protocol is preferential to use of the random access protocol, since the reserved access protocol uses the channel resources with greater efficiency. When the mobile station has one or more bursts to transmit in CoS2 for which it does not have a reservation, it shall transmit according to the random access procedure defined in Section 5.6.2, using the random access parameters defined in Section Note. The random access procedure uses a non-adaptive p-persistent algorithm Random access parameters The random access protocol shall implement the system parameters defined in Table 20. Symbol Parameter name Minimum Maximum Recommended default Increment p1 Persistence CoS2 1/ /64 1/64 VS3 Maximum number of access attempts Table 23: Random access parameters for CoS Parameter p1 (Persistence CoS2) Parameter p1 (Persistence CoS2) shall be the probability that the station will transmit any random access attempt in CoS Counter VS3 (maximum number of access attempts) Counter VS3 shall be used to limit the maximum number of random access attempts (VS3) that a station will make for any transmission request in CoS2. The VS3 counter shall be cleared upon system initialization, reaching the end of the CoS2 section, or a successful access attempt. The VS3 counter shall be incremented after every unsuccessful random access attempt in CoS2. When the VS3 counter reaches the maximum number of random access attempts, authorization to transmit shall be granted as soon as the channel is available. Edition: 1.0 Draft Page 63

64 5.6.2 Random access procedures When the station has one or more bursts to transmit for which it does not have a reservation, it shall use the p-persistent algorithm defined as follows: a) The station shall select a slot, or block of slots as required, for the data in its random access queue by choosing a slot or block of slots that have not been reserved by the ground station for other mobile stations. b) If the station is able to select a slot or block of slots, then the station shall transmit in the slot, or block of slots with probability p1 (defined in ) c) The station shall clear the VS3 counter (VS3 counter cancelled) if it was able to transmit. d) If the mobile station is unable to transmit at the first attempt, and it has not reached the end of the current CoS2 section, and the VS3 timer has not expired, it shall increment the VS3 counter, and then make a further random access attempt, restarting the process at a). e) If the mobile station is unable to select a slot at the first attempt, and it has not reached the end of the current CoS2 section, and the VS3 timer has expired, it shall select slots as in a) and then transmit in those slots with probability p1=1. If the mobile station is unable to select a slot within the current CoS2 section, this shall be regarded as an unsuccessful random access attempt, and the mobile station shall issue an RTS request for slots to send the data in its next CoS1 slot, and clear the VS3 counter (VS3 counter cancelled). 5.7 Random access protocol specification for transmission in LoG2 When the mobile station needs to login to a GS, it shall transmit in the LoG2 section of the frame according to the random access procedure defined in Section 5.7.2, using the random access parameters defined in Section Note. The random access procedure uses a non-adaptive p-persistent algorithm Random access parameters The random access protocol shall implement the system parameters defined in Table 21. Symbol Parameter name Minimum Maximum Recommended default Increment p2 Persistence LoG2 1/ /64 1/64 VS4 Maximum number of access attempts Table 24: Random access parameters for LoG Parameter p2 (Persistence LoG2) Parameter p2 (Persistence LoG2) shall be the probability that the station will transmit any random access attempt in LoG Counter VS4 (maximum number of access attempts) Counter VS4 shall be used to limit the maximum number of random access attempts (VS4) that a station will make for any transmission request in LoG2. The VS4 counter shall be cleared upon system initialization, reaching the end of the LoG2 section, or a successful access attempt. The VS4 counter shall be incremented after every unsuccessful random access attempt in Edition: 1.0 Draft Page 64

65 LoG2. L-DACS2 System Definition proposal: Deliverable D2 When the VS4 counter reaches the maximum number of random access attempts, authorization to transmit shall be granted as soon as the channel is available Random access procedures When the station has a login bursts to transmit, it shall use the p-persistent algorithm defined as follows: a) The station shall select a slot at random from the available slots in the LoG2 section, such that it is equally likely to select any of the available slots. b) If the station is able to select a slot, then the station shall transmit in the slot with probability p2 (defined in ). c) The station shall clear the VS4 counter (VS4 counter cancelled) if it was able to transmit. d) If the mobile station is unable to transmit at the first attempt, and it has not reached the end of the current LoG2 section, and the VS4 timer has not expired, it shall increment the VS4 counter, and then make a further random access attempt, restarting the process at a). e) If the mobile station is unable to select a slot at the first attempt, and it has not reached the end of the current LoG2 section, and the VS4 timer has expired, it shall select slots as in a) and then transmit in those slots with probability p=1. If the mobile station is unable to select a slot within the current LoG2 section, this shall be regarded as an unsuccessful random access attempt, and the mobile station shall clear the VS4 counter (VS4 counter cancelled) and restart the process for the LoG2 section in the following frame. Edition: 1.0 Draft Page 65

66 CHAPTER 6 L-DACS2 Link Management layer (LML) 6.1 Introduction Each ground station and each aircraft mobile station supporting air/ground point-to-point communication services shall include the functionality of an L-DACS2 Link Management Entity (LLM). An LLM shall be responsible for the data link management policy of the system. In a mobile system, the LLM shall be responsible for determining with which ground station(s) the aircraft system should maintain data link communication at any given time. In a ground station, the LLM shall be responsible for determining which aircraft mobile system(s) should be provided with data link communications. An LLM shall have a Link Management Entity (LME) for each peer LME. Hence, a ground station LLM shall have an LME for each aircraft mobile system and an aircraft mobile LLM shall have an LME for each ground station with which it is communicating. Note. If an aircraft s mobile system receives a burst from a ground station, only one LME will process and react to that burst. The L-DACS2 link management layer (LML) is divided into four sub-layers: Media Access Control (MAC) sublayer that requires the use of Time Division Multiple Access (TDMA); an L-DACS2 Services sub-layer (LSS) that provides communication by using a flexible burst format and associated transmission and reservation protocols over the MAC sub-layer; a Data Link Services sublayer (DLS) that provides connection-oriented and broadcast services over the LSS; a Link Management Entity (LME) that establishes and maintains connections. The LML shall provide a reliable point-to-point service by using a connection-orientated DLS. The LML shall provide an unacknowledged broadcast service by using a connectionless Edition: 1.0 Draft Page 66

67 DLS. L-DACS2 System Definition proposal: Deliverable D2 Note. There is one LSS entity for each L-DACS2 channel that is accessed by the station (airborne or ground). The LSS provides service to the LLM as well as to the LME associated with other L-DACS2 peer systems, their associated Data Link Entities (DLEs) and the DLS. The LSS is served by the MAC that is associated with its particular L-DACS2 channel. 6.2 Login mechanism An aircraft entering a cell shall know (from on-board information) the frequency of the corresponding ground station. The aircraft shall listen on this frequency for the framing message transmitted by the ground station. Note. This message contains information about the slot structure. The aircraft shall randomly select one of the login slots in the LoG2 section of the frame. It shall then announce its presence to the ground station by transmitting a message in the chosen login slot. Note. Randomly selecting a login slot reduces the likelihood of the aircraft attempting to login at the same time in the same slot as another aircraft. Note. The aircraft is expected to transmit this login message in the same frame as the framing message which it received (i.e. less than 1 second later). The ground station shall reply in UP1 or UP2 informing the aircraft of the position of its allocated slot in the CoS1 section. The ground station shall give the aircraft a local 9-bit address. The ground station shall inform the aircraft of the ground station s own local (7-bit) address. Note. These short addresses are used within the cell for identification instead of the longer 27-bit ICAO address. The aircraft shall then be able to transmit RTSs in its allocated slot in the CoS1 section of the same frame. Note. It is expected that the aircraft will be able to transmit in an allocated CoS1 slot very soon after reaching the new cell (as a framing message is transmitted by the ground station every 1 second in the default setting). 6.3 Hand-over mechanism There shall be two possible means of hand-over: controlled and uncontrolled. Note. Hand-over principles are very similar in various types of systems 10. Note. The L-DACS2 hand-over procedure differs from the GSM hand-over procedure because it is expected that most hand-overs will be mobile-initiated. Ground stations may request hand-overs but cannot initiate them. This is to avoid hang-ups. Passive scanning at the mobile is used to detect nearby ground stations according to the local link management policy. Note. With current technology this may require separate receivers since the aircraft needs to be monitoring the UP segment in the current channel all the time. However there are recent trends in development of software radio capability that allow software functions to scan other parts of the band that the radio is not tuned into Edition: 1.0 Draft Page 67

68 6.3.1 Controlled hand-over Controlled hand-overs can be air-initiated or ground-requested air-initiated. Note. A controlled hand-over is a make-before-break ( soft ) hand-over Air-initiated controlled hand-over An air-initiated controlled hand-over shall be triggered if any of the following events occur: According to the local link management policy, the signal quality on the current link is deemed to be insufficient for maintaining reliable communications and the signal quality of another ground station is significantly better, for more than the channel timeout period. The channel busy timer expires. In this case, the mobile LME shall only initiate the handover procedure if the signal from another ground station is of sufficient signal quality. The mobile LME is at a position which, according to the local link management policy, requires that a connection with a new ground station is established. Note. An aircraft shall know, from on-board information, when it is nearing the edge of the coverage area of the current cell. Note. If an aircraft has commenced approach to its destination airport and its current link is with a ground station that does not offer data link service at that airport, then the aircraft should hand-over as soon as possible to a ground station that does offer data link service at that airport. At an appropriate time, the aircraft shall transmit a cell exit message to the ground station in CoS1. Note. This appropriate time shall be implementation-dependent. It may be influenced by factors such as received signal power (SQP) and bit error rate (BER). As the cell exit message is being transmitted, the aircraft shall also commence the login procedure described in section 6.2. Note. The aircraft will know which frequency to search from its on-board information. The assumption is that the aircraft knows its own position and the positions of all ground stations and their frequencies. Prediction of which ground stations should be connected to base on aircraft route tracing and prediction is not assumed to be required, but may improve handover reliability. The current ground station shall reply to the aircraft s cell exit message in an UP1 or UP2 slot, to confirm that the aircraft is leaving the cell When the aircraft receives the exit confirmation message from the current ground station in the UP1 or UP2 slot, it shall send an ACK message to the ground station, indicating that the exit confirmation message is being acknowledged. Note. This will be the last message that the aircraft sends to the current ground station. Note. The aircraft will not do this until is has correctly logged-on to the next ground station. If the connection with the current ground station is lost before the aircraft has acknowledged the exit confirmation message, then the slot will be de-allocated according to the procedure in section When the current ground station receives the ACK message from the aircraft, it shall deallocate the aircraft s CoS1 slot and shall consider the link to be terminated. See Figure 12 for the diagram of a successful handover. Although the aircraft may simultaneously be in communication with both the current and next ground stations, no data messages shall be transmitted whilst the hand-over process is occurring. Data messages shall only be sent once the hand-over is completed or failed. Edition: 1.0 Draft Page 68

69 Previous ground station AC Cell exit [COS1 slot] GS_EXIT_ACK Ack = 1 [UP1 or UP2] Aircraft Listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Next ground station Address allocated Slot allocated [UP1 or UP2] *AC_ACK Slot de-allocated Hand-over complete Ack = 1, ID = CELL_EXIT [COS2] *Only performed if still in range. If AC_ACK is not transmitted, the link will time-out Figure 12: Successful air-initiated controlled hand-over Previous ground station AC Cell exit [COS1 slot] GS_EXIT_ACK Ack = 1 [UP1 or UP2] Aircraft Listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] Next ground station Not received correctly Next frame retransmit CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] Slot de-allocated Hand-over complete *AC_ACK Ack = 1, ID = CELL_EXIT [COS2] *Only performed if still in range. If AC_ACK is not transmitted, the link will time-out Figure 13: Air-initiated controlled handover: retransmit cell login Edition: 1.0 Draft Page 69

70 Previous ground station AC Cell exit [COS1 slot] GS_EXIT_ACK Ack = 1 [UP1 or UP2] Aircraft Listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Next ground station Address allocated Slot allocated Not received correctly [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC New slot allocated (Allocated address is re-used) [UP1 or UP2] Slot de-allocated Hand-over complete *AC_ACK Ack = 1, ID = CELL_EXIT [COS2] *Only performed if still in range. If AC_ACK is not transmitted, the link will time-out Figure 14: Air-initiated controlled handover: retransmit GS_ALLOC Not received correctly Previous ground station AC Cell exit [COS1 slot] GS_EXIT_ACK Ack = 0 [UP1 or UP2] AC Cell exit [COS1 slot] GS_EXIT_ACK Aircraft Listen on appropriate frequency GS_FRAME *Next frame retransmit CELL_LOGIN [LoG2 slot] GS_ALLOC [UP1 or UP2] Next ground station Address allocated Slot allocated Ack = 1 [UP1 or UP2] *AC_ACK Slot de-allocated Hand-over complete Ack = 1, ID = CELL_EXIT [COS2] *Only performed if still in range, otherwise link will time-out If AC_ACK is not transmitted, the link will time-out Figure 15: Air-initiated controlled handover: retransmit AC_CELL_EXIT Edition: 1.0 Draft Page 70

71 Previous ground station AC Cell exit [COS1 slot] Aircraft Listen on appropriate frequency GS_FRAME Next ground station GS_EXIT_ACK Ack = 1 [UP1 or UP2] AC_ACK Not received correctly CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated Ack = 0, ID = CELL_EXIT [COS2] [UP1 or UP2] AC Cell exit [COS1 slot] *Next frame retransmit GS_EXIT_ACK Ack = 1 [UP1 or UP2] Slot de-allocated Hand-over complete *AC_ACK Ack = 1, ID = CELL_EXIT [COS2] *Only performed if still in range, otherwise link will time-out If AC_ACK is not transmitted, the link will time-out Figure 16: Air-initiated controlled handover: retransmit GS_EXIT_ACK Ground-requested air-initiated controlled hand-over The ground station may be able to determine, from the location information in an aircraft s ADS-B transmissions, when the aircraft is nearing the edge of a cell. A ground-requested air-initiated controlled hand-over may be triggered if any of the following events occurs: According to the local link management policy, the current ground station determines that the aircraft is leaving the cell and is sufficiently close to a neighbouring ground station for the hand-over process to commence. According to the local link management policy, the current ground station determines that the aircraft is sufficiently close to a neighbouring ground station for the hand-over process to commence and the current ground station data load has exceeded the overload setting. Note. The overload setting for a ground station is implementation-dependent. Note. The overload hand-over process attempts to share channel loading between adjacent ground stations. The signal quality from the aircraft may not be sampled to trigger a ground-requested airinitiated hand-over. If the ground station receives garbled data it shall only reply with a NACK. Note. If the signal quality has degraded to an unacceptable level then the aircraft will initiate the hand-over process. The current ground station shall have knowledge of the cell loadings of the other ground stations in its cluster from the GNI. If it is possible for the aircraft to hand-over to a selection of ground stations then the current ground station shall compare the cell loadings of these ground stations. If the ground station determines that a hand-over is appropriate, it shall transmit a cell exit message to the aircraft. If the current ground station determines that the aircraft should handover to a particular neighbouring ground station then this shall be indicated in the cell exit Edition: 1.0 Draft Page 71

72 message. L-DACS2 System Definition proposal: Deliverable D2 Note. The criteria for determining whether or not a hand-over is appropriate are implementation specific. Note. The indication of which neighbouring ground station the aircraft should contact is optional, is only possible within a GNI cluster, and is dependent on knowledge of adjacent ground stations loadings. The aircraft, on receiving a cell exit message from the current ground station, shall attempt communication with the ground station in the next cell. If the cell exit message indicates a particular future ground station then the aircraft shall preferentially attempt to communicate with it (see Figure 22). If the signal quality from the suggested future ground station is too low then the aircraft shall attempt communication with any other possible future ground station (see Figure 23). Note. The aircraft will know which frequency to search for from its on-board information. Because both the aircraft and the current ground station know the location and frequency channel of the ground station in the next cell, it is not necessary for the current ground station to transmit any information other than a cell exit trigger. This differs from the typical GSM/GPRS procedure. If the communication with the next ground station is successful and the aircraft is allocated a CoS1 slot in the next cell, it shall reply to the current ground station with an exit confirmation message indicating that the current link can be terminated (see Figure 17). If the communication with the next ground station is not successful, the aircraft shall reply to the current ground station with an exit confirmation message indicating that the current link cannot be terminated (see Figure 20).. When the current ground station receives an exit confirmation message showing that the current link can be terminated, it shall transmit an ACK to the aircraft, shall de-allocate the aircraft s CoS1 slot and shall consider the link to be terminated (see Figure 17). If the current ground station receives an exit confirmation message from the aircraft showing that the current link cannot be terminated, and it still desires that the hand-over is carried out, then it shall re-transmit the cell exit message to the aircraft in the next frame (see Figure 21). Otherwise the aircraft shall maintain its communication with the current ground station. The ground-requested air-initiated controlled hand-over from the current ground station shall not be completed before the aircraft has successfully made contact with the next cell. Note. If contact between the aircraft and the current ground station is lost before the ground-requested air-initiated controlled hand-over has been completed, then the uncontrolled hand-over process will be initiated. Note. Unlike Figure 13 to Figure 16, not all possible login error conditions are assessed in the following diagrams; the process in Figure 21 will take place if a login is unsuccessful for any reason. Edition: 1.0 Draft Page 72

73 Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft starts to listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] AC_EXIT_ACK Slot de-allocated Ack = 1 [COS1 slot] GS_ACK Ack = 1, ID = EXIT_ACK [UP1 or UP2] Hand-over complete If not received, then aircraft still considers exit to be complete Figure 17: Successful ground-requested air-initiated controlled hand-over Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] AC_EXIT_ACK Not received correctly Ack = 0, Ack slot number = 0 [COS1 slot] Next frame retransmit GS_CELL_EXIT [UP1 or UP2] Aircraft starts to listen on appropriate frequency GS_FRAME [UP1 or UP2] [Repeat as Figure 15] Figure 18: Ground-requested air-initiated controlled hand-over: GS_CELL_EXIT retransmit Edition: 1.0 Draft Page 73

74 Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft starts to listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] AC_EXIT_ACK Not received correctly Ack = 1 [COS1 slot] GS_ACK Ack = 0, ID = EXIT_ACK [UP1 or UP2] AC_EXIT_ACK Ack = 1 [COS1 slot] GS_ACK Hand-over complete Ack = 1 ID = EXIT_ACK [UP1 or UP2] Figure 19: Ground-requested air-initiated controlled hand-over: AC_EXIT_ACK retransmit Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft starts to listen on appropriate frequency AC_EXIT_ACK GS_FRAME not received. Aircraft notes login as unsuccessful Ack = 0, Ack slot number = 0 [COS1 slot] *Next frame retransmit GS_CELL_EXIT [UP1 or UP2] [Repeat as Figure ] *Only performed if aircraft is still in range, otherwise link will time-out Figure 20: Ground-requested air-initiated controlled hand-over: GS_FRAME not received Edition: 1.0 Draft Page 74

75 Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft starts to listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN AC_EXIT_ACK Aircraft notes login as unsuccessful [LoG2 slot] Slot not allocated (for any reason) Ack = 0, Ack slot number = 0 [COS1 slot] *Next frame retransmit GS_CELL_EXIT [UP1 or UP2] [Repeat as Figure 15] *Only performed if aircraft is still in range, otherwise link will time-out Figure 21: Ground-requested air-initiated controlled hand-over: unsuccessful login Previous ground station GS_CELL_EXIT Aircraft Recommended ground station [UP1 or UP2] Aircraft starts to listen on recommended frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] AC_EXIT_ACK Slot de-allocated Ack = 1 [COS1 slot] GS_ACK Ack = 1, ID = EXIT_ACK [UP1 or UP2] Hand-over complete If not received, the aircraft still considers exit to be complete Figure 22: Successful ground-requested air-initiated controlled hand-over with recommendation Edition: 1.0 Draft Page 75

76 Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft does not recognize the recommended GS, or the login procedure is unsuccessful Aircraft determines another possible GS Aircraft starts to listen on appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] AC_EXIT_ACK Slot de-allocated Ack = 1 [COS1 slot] GS_ACK Ack = 1, ID = EXIT_ACK [UP1 or UP2] Hand-over complete If not received, the aircraft still considers exit to be complete Figure 23: Ground-requested air-initiated controlled hand-over with recommendation: alternative ground station Previous ground station GS_CELL_EXIT Aircraft Next ground station [UP1 or UP2] Aircraft starts to listen on appropriate frequency Login attempts with recommended GS and other possible GSs are all unsuccessful AC_EXIT_ACK Aircraft notes login as unsuccessful Ack = 0, Ack slot number = 0 [COS1 slot] *Next frame retransmit GS_CELL_EXIT [UP1 or UP2] [Repeat as Figure 20] *Only performed if aircraft is still in range, otherwise link will time-out Figure 24: Ground-requested air-initiated controlled hand-over with recommendation: no successful logins Edition: 1.0 Draft Page 76

77 6.3.2 Uncontrolled hand-over If communication between the aircraft and the ground station is lost, for more than a predetermined time period, then both the aircraft and the ground station shall consider their link to be terminated. This process could be assured by the monitoring of the aircraft traffic. Indeed as described in section 4.5.1, the mobile shall send at least a keep alive message in a specific time frame defined by S1 parameter. Note. The purpose of the time-out period is to avoid brief signal fluctuations causing unnecessary hand-over procedures. If the signal is not detected for this significant period of time then it is likely that the aircraft has moved out of the relevant ground station s coverage. Note. An uncontrolled hand-over may be a break-before-make ( hard ) hand-over if the aircraft does not successfully login with another ground station before considering the connection with the previous one to be lost. Note. The appropriate time-out period shall be implementation-dependent. For this process to occur correctly, the appropriate value for the time-out period must be chosen (as it may be affected by local factors). The ground station shall assume that the aircraft s dedicated CoS1 slot is no longer required by the aircraft and shall de-allocate it. The aircraft shall determine the appropriate (new) ground station to contact and shall begin the login procedure on the relevant frequency. Note. This procedure means that de-allocation of the old link may occur before the new link has been established. This is a hard hand-over. The process is illustrated in Figure 25; note that it is assumed in this diagram that the receipts of the GS_FRAME, CELL_LOGIN and GS_ALLOC messages are successful. For alternative situations during login see Figure 13 to Figure 16. Previous ground station Aircraft Next ground station [Link time-out occurs] Slot de-allocated Aircraft considers link to be lost Aircraft determines possible GS to contact and starts to listen on the appropriate frequency GS_FRAME [UP1 or UP2] CELL_LOGIN [LoG2 slot] GS_ALLOC Address allocated Slot allocated [UP1 or UP2] [Repeat as for Figures 10-14] Figure 25: Uncontrolled handover Edition: 1.0 Draft Page 77

78 ANNEX 1 Example of burst transmission A1.1 Burst transmission in UP1 or UP2 Table 12 derives the number of user octets in a burst designed to fit in two consecutive slots in UP1 or UP2. This is just an example and on average it expected that a burst in UP1 or UP2 will occupy several or tens of slots, in which case the guard time (which is larger than for e.g. CoS2) will be proportionally smaller when compared to the amount of user data transmitted per unit of slot length. Edition: 1.0 Draft Page 78

79 Parameter description Derivation Value Duration of 1 symbol/bit (T S ) Number of symbol durations (T S ) per frame Duration of UP1/UP2 slot (M5) 6/ s or ~ ms /6 bps or ~ bps 1/150 s or ~ ms Bits per 1 slot /(6 * 150) bits /9 bits or ~ bits Bits per 2 slots 2 * /(6 * 150) bits /9 bits or ~ bits Guard time prior to start of burst See Section bits or ~ ms Guard time at end of burst See Table 5 in Section /9 bits or bits ~ ms Active slot duration Bits per slot start guard time bits end guard time bits 2448 bits Transmitter stabilisation Ramp up time + ramp down time 16 bits Usable slot duration Bits per slot (ramp up bits + ramp down bits) guard time bits 2432 bits Synchronisation (training) sequence Based on GSM ( bits) 26 bits Start and end flags 2 x 8 bits 16 bits CRC 16 bits 16 bits Bits available for FEC, addresses and user data Bits available after deduction of bits required for FEC Octets available after deduction of FEC overhead Approximate number of octets required for: version number, source address, message type, message ID, destination address, data length Approximate number of octets available for user data Bits per active slot 58 bits 30% FEC coding: divide bits available by 1.3, round down to nearest octet 2374 bits Edition: 1.0 Draft Page Bits available/ bit address (includes 24 bit ICAO address see Note 1); 2-bit version number 4 octets 224 octets Use of ICAO 24-bit address is worst case. Mobile stations and ground stations will be allocated local addresses by the ground station, which will save on bits. Table 25: Available payload for user data in burst occupying two slots in UP1 or UP2

80 The following figure illustrates the number of bits allocated to each section of a burst transmitted in the UP1 or UP2 section. For the purposes of illustration, this burst occupies two basic (full) slots, each of ms duration. In practice, the uplink transmissions can be as long as the entire UPx section, depending on how much data is scheduled for transmission. Burst Occupying two ms slots ( bits) Start guard time Ramp up Sync sequence Start flag User data Addresses, msg IDs User data Application message 408 bits 8 bits 26 bits 8 bits 32 bits (4 octets) 1792 bits (224 octets) FEC CRC End flag Ramp down End guard time 550 bits 16 bits 8 bits 8 bits bits Figure 26: Illustration of burst format for a typical burst in UP1 or UP Burst transmission in a single slot in CoS2 Table 13 derives the number of user octets in a burst designed to fit in a single slot in CoS2. 11 Note that in this illustrative example, there is only one message in this uplink burst, hence only one FEC field. Edition: 1.0 Draft Page 80

81 Parameter description Derivation Value Duration of 1 symbol/bit (T S ) Number of symbol durations (T S ) per frame Duration of CoS2 slot (M5) 6/ s or ~ ms /6 bps or ~ bps 1/150 s or ~ ms Bits per slot /(6 * 150) bits /9 bits or ~ bits Guard time Reduced guard time based on use of Timing Advance (see Section ) /9 bits or bits ~ ms Active slot duration Bits per slot guard time bits 1772 bits Transmitter stabilisation Ramp up time + ramp down time 16 bits Usable slot duration Bits per slot (ramp up bits + ramp down bits) guard time bits 1756 bits Synchronisation (training) sequence Based on GSM ( bits) 26 bits Start and end flags 2 x 8 bits 16 bits CRC 16 bits 16 bits Bits available for FEC, addresses and user data Bits available after deduction of bits required for FEC Octets available after deduction of FEC overhead Approximate number of octets required for: version number, source address, message type, message ID, ground station local address, acknowledgement of uplinked data, data length, Request-To-Send (RTS) Approximate number of octets available for user data Bits per active slot 58 bits 30% FEC coding: divide bits available by 1.3, round down to nearest octet 1698 bits 1304 Bits available/ bit address (includes 24 bit ICAO address see Note 1); 2-bit version number 8 octets 155 octets Use of ICAO 24-bit address is worst case. Mobile stations and ground stations will be allocated local addresses by the ground station, which will save on bits. Table 26: Available payload for user data in a single slot in CoS2 The following figure illustrates the number of bits allocated to each section of a single slot burst in CoS2. Edition: 1.0 Draft Page 81

82 Full size slot length ms ( bits) Ramp up Sync sequence Start flag User data Addresses, msg IDs User data Application message FEC CRC End flag Ramp down Guard time 8 bits 26 bits 8 bits 64 bits (8 octets) 1240 bits (155 octets) 394 bits 16 bits 8 bits 8 bits bits Figure 27: Illustration of burst format for a single slot burst in CoS2 6.5 Burst transmission in CoS1 slot Table 14 derives the number of user octets in a burst designed to fit in a single CoS1 slot. It is currently not intended that a CoS1 slot carries data messages, apart from ACK and RTS. Therefore this table calculation serves as a check that the CoS1 slot size is sufficient to accommodate those message fields that are required. Edition: 1.0 Draft Page 82

83 Parameter description Derivation Value Duration of 1 symbol/bit (T S ) Number of symbol durations (T S ) per frame Duration of CoS1 slot (M7) 6/ s or ~ ms /6 bps or ~ bps 1/900 s or ~ ms Bits per CoS1 slot /(6 * 900) bits /27 bits or ~ bits Guard time Reduced guard time based on use of Timing Advance (see Section ) /27 bits or bits ~ ms Active slot duration Bits per slot guard time bits 267 bits Transmitter stabilisation Ramp up time + ramp down time 16 bits Usable slot duration Bits per slot (ramp up bits + ramp down bits) guard time bits 251 bits Synchronisation (training) sequence Based on GSM ( bits) 26 bits Start and end flags 2 x 8 bits 16 bits CRC 16 bits 16 bits Bits available for FEC, addresses and user data Bits available after deduction of bits required for FEC Octets available after deduction of FEC overhead Approximate number of octets required for: version number, source address, message type, message ID, ground station local address, acknowledgement of uplinked data, data length, Request-To-Send (RTS) Approximate number of octets available for user data Bits per active slot 58 bits 30% FEC coding: divide bits available by 1.3, round down to nearest octet 193 bits Edition: 1.0 Draft Page Bits available/ bit address (includes 24 bit ICAO address see Note 1); 2-bit version number 7 octets 11 octets Use of ICAO 24-bit address is worst case. Mobile stations and ground stations will be allocated local addresses by the ground station, which will save on bits. Table 27: Available payload for user data in a CoS1 slot As an option for future message definition, the 11 octets available for user data could be used to transmit the position of the aircraft. This information could be used for independent/inherent security functions, surveillance functions, communication system management etc. This would assume that position data is available from an onboard

84 navigation system. Further analysis should determine if the inclusion of position information would provide a benefit. The following figure illustrates the number of bits allocated to each section of a burst in CoS1. CoS1 slot length ms ( bits) Ramp up Sync sequence Start flag User data Addresses, msg IDs User data Application message FEC CRC End flag Ramp down Guard time 8 bits 26 bits 8 bits 56 bits (7 octets) 88 bits (11 octets) 49 bits 16 bits 8 bits 8 bits bits Figure 28: Illustration of burst format for a burst in a CoS1 slot 6.6 Burst transmission in a LoG2 slot Table 15 derives the number of user octets in a burst designed to fit in a LoG2 slot. It is currently not intended that a LoG2 slot caries data messages. Therefore this table calculation serves as a check that the LoG2 slot size is sufficient to accommodate those message fields that are required. Edition: 1.0 Draft Page 84

85 Parameter description Derivation Value Duration of 1 symbol/bit (T S ) Number of symbol durations (T S ) per frame Duration of LoG2 slot (M6) 6/ s or ~ ms /6 bps or ~ bps 1/300 s or ~ ms Bits per LoG2 slot /(6 * 300) bits /9 bits or ~ bits Guard time See Section /9 bits or bits ~ ms Active slot duration Bits per slot guard time bits 557 bits Transmitter stabilisation Ramp up time + ramp down time 16 bits Usable slot duration Bits per slot (ramp up bits + ramp down bits) guard time bits 541 bits Synchronisation (training) sequence Based on GSM ( bits) 26 bits Start and end flags 2 x 8 bits 16 bits CRC 16 bits 16 bits Bits available for FEC, addresses and user data Bits available after deduction of bits required for FEC Octets available after deduction of FEC overhead Approximate number of octets required for: version number, ICAO 27-bit source address, message type, message ID, ground station 27-bit ICAO address, authentication Approximate number of octets available for additional user data (if required) Bits per active slot 58 bits 30% FEC coding: divide bits available by 1.3, round down to nearest octet 483 bits 368 Bits available/ bit address (includes 24 bit ICAO address); 2-bit version number Table 28: Available payload for user data in a LoG2 slot 13 octets 33 octets The following figure illustrates the number of bits allocated to each section of a burst in a LoG2 slot. Edition: 1.0 Draft Page 85

86 LoG2 slot length ms ( bits) Ramp up Sync sequence Start flag User data Addresses, msg IDs User data Application message FEC CRC End flag Ramp down Guard time 8 bits 26 bits 8 bits 104 bits (13 octets) 264 bits (33 octets) 115 bits 16 bits 8 bits 8 bits bits Figure 29: Illustration of burst format for a burst in a LoG2 slot Edition: 1.0 Draft Page 86

87 ANNEX 2 Message structure A2.1 Message type codes This section shows diagrams for all of the message types, showing the fields and the numbers of bits required for each one. The following assumptions have been made: All messages shall start with the ISO/IEC flag Single-bit flags are used for ACK, RTS and CTS ICAO 27-bit addresses are only used when necessary Aircraft local addresses shall be 9-bit addresses is not permitted means broadcast to all aircraft and is only used by the GS Message type is followed by the destination address the station will analyse the source address and the message type to see if the message is expected A2.2 Message type codes 6 bits are used for message types to allow space for future codes is not used. All binary codes which are not shown are unallocated. Edition: 1.0 Draft Page 87

88 Message type Message name Binary code Mobile station CoS1 RTS/ACK downlink RTS_COS Mobile station CoS1 Keep alive KEEP_ALIVE Mobile station CoS2 data downlink DATA_COS Mobile station CoS2 random access data downlink DATA_RA_COS Mobile station cell exit AC_CELL_EXIT Ground station cell exit GS_CELL_EXIT Mobile station cell exit ACK AC_EXIT_ACK Ground station cell exit ACK GS_EXIT_ACK Ground station data uplink GS_DATA Ground station framing message GS_FRAME Ground station ACK GS_ACK Mobile station CoS2 ACK AC_ACK Ground station ACK/CTS ALL message CTS_ACK_ALL Mobile station LoG2 login CELL_LOGIN Ground station login accept GS_ALLOC Grand station timing advance GS_TADVANCE Table A-1 : Message type codes A2.3 Priority field The priority field follows the ATN numbering scheme and can have values from 0 to 14. A2.4 Messages Note. Power control fields have to be developed in the uplink message structure. This is under development. Note. The CRC has been estimated to be sufficient at 2 octets per message. This will be validated in a next version of the document. Edition: 1.0 Draft Page 88

89 Login message: CELL_LOGIN This message shall be transmitted by a mobile station in a LoG2 slot. Note. 384 bits are available for user data in a LoG2 slot. ISO flag 8 Binary Version number 2 Binary 00 Address length flag 1 Binary 1 for 27-bit ICAO address A/C ICAO address 27 Message type 6 Binary Message identifier 6 1 to 64 ( to ) GS ICAO address 27 Destination ground station Authentication 32 Size not fixed FEC 31 30% FEC coding CRC 16 ISO flag 8 Binary bits The mobile station will listen for the framing message to identify the position in the frame of the LoG2 slots. The ground station reply to the login message will be transmitted in the UP1 section of the following frame. Login reply message: GS_ALLOC This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Binary Version number 2 Binary 00 Address length flag 1 Binary 1 GS ICAO address 27 Message type 6 Binary Login message identifier 6 1 to 64 ( to ) A/C ICAO address 27 A/C local address possible addresses (all 0 s and all 1 s invalid) GS local address possible addresses (all 0 s and all 1 s invalid) CoS1 slot number 9 Measured from start of frame, range Power Control Field for not monitor, range 1-25 FEC 29 30% FEC coding CRC 16 ISO flag 8 Binary bits The ground station will transmit a reply to the mobile station in UP1 after reception of a login message from that mobile station. The allocated CoS1 slot will be exclusively assigned to the mobile station until that mobile station leaves the cell. Edition: 1.0 Draft Page 89

90 Cell exit message: AC_CELL_EXIT This message shall be transmitted by a mobile station in CoS1. Note. With M4 = 6, 235 bits are available for this message in a CoS1 slot. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 1 ICAO address 27 Transmitting A/C Message type 6 Binary Message identifier 6 1 to 64 ( to ) ICAO address 27 Current ground station Next station flag 1 Binary 0 or 1 Next station address 27 Suggested future ground station Authentication 32 Size not fixed FEC 39 30% FEC coding CRC 16 ISO flag 8 Binary bits If the next station flag is 0 then the next station address field shall be all 0's. Cell exit message: GS_CELL_EXIT This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 1 ICAO address 27 Transmitting GS Message type 6 Binary Message identifier 6 1 to 64 ( to ) ICAO address 27 Destination A/C Next station flag 1 Binary 0 or 1 Next station address 27 Suggested future ground station Authentication 32 Size not fixed FEC 39 30% FEC coding CRC 16 ISO flag 8 Binary bits If the next station flag is 0 then the next station address field shall be all 0's. Edition: 1.0 Draft Page 90

91 Cell exit reply message: AC_EXIT_ACK This message shall be transmitted by a mobile station in CoS2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 1 ICAO address 27 Transmitting A/C Message type 6 Binary Exit message identifier 6 1 to 64 ( to ) ICAO address 27 Destination GS ACK flag 1 ACK slot number 9 CoS2 slot for the A/C ACK, range Set to 0 for an A/C reply to a GS CELL_EXIT FEC 24 30% FEC coding CRC 16 ISO flag 8 Binary bits Cell exit reply message: GS_EXIT_ACK This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 1 ICAO address 27 Transmitting GS Message type 6 Binary Exit message identifier 6 1 to 64 ( to ) ICAO address 27 Destination A/C ACK flag 1 ACK slot number 9 CoS2 slot for the A/C ACK, range Set to 0 for an A/C reply to a GS CELL_EXIT FEC 24 30% FEC coding CRC 16 ISO flag 8 Binary bits Edition: 1.0 Draft Page 91

92 Framing message: GS_FRAME L-DACS2 System Definition proposal: Deliverable D2 This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 1 GS ICAO address 27 Necessary for aircraft entering the cell to register this message Message type 6 Binary Message identifier 6 1 to 64 ( to ) A/C local address 9 Binary to indicate broadcast UTC time and date 28 Used for synchronization and time stamping Active UP1 length 7 Current number of slots, Active CoS1 length 9 Current number of slots, Active UP2 length 7 Current number of slots, Active CoS2 length 8 Current number of slots, Active LoG2 length 4 Current number of slots, 4-33 New frame number 8 Frame which will change to new section sizes, 0 if section sizes are not changing. New UP1 length 8 New number of slots in UP1, New CoS1 length 8 New number of slots in CoS1, New UP2 length 8 New number of slots in UP2, New CoS2 length 8 New number of slots in CoS2, New LoG2 length 4 New number of slots in LoG2, 4-33 FEC 48 30% FEC coding CRC 16 ISO flag 8 Binary bits The following fields: New UP1 length New CoS1 length New UP2 length New CoS2 length New LoG2 length shall be ignored by the mobile station if the New frame number field is set to 0. If the mobile station receives a GS_FRAME message in which some or all of the section sizes are out-of-range, then it shall transmit a NACK to the GS and use the existing frame section sizes. Edition: 1.0 Draft Page 92

93 GS data uplink message: GS_DATA This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 GS local address 7 Message type 6 Binary Message identifier 6 1 to 64 ( to ) A/C local address 9 Power Control Field for not monitor, range 1-25 Data length (octets) 11 Range 1-2,048 octets Data Variable length FEC % FEC coding CRC 16 ISO flag 8 Binary bits CoS1 keep-alive message: KEEP_ALIVE This message shall be transmitted by a mobile station in CoS1. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 A/C local address 9 Message type 6 Binary Message identifier 6 1 to 64 ( to ) GS local address 7 FEC 10 30% FEC coding CRC 16 ISO flag 8 Binary bits This message is transmitted by the mobile station when it has no data and no acknowledgement to send. Edition: 1.0 Draft Page 93

94 CoS1 RTS downlink message: RTS_COS1 This message shall be transmitted in CoS1 by a mobile station that has a requirement to transmit an ACK and/or an RTS to the ground station. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 A/C local address 9 Message type 6 Binary Message identifier 6 1 to 64 ( to ) GS local address 7 ACK flag 1 Reply to UP1 uplink ACK of message type 6 Message type acknowledged (all 0's if unknown) ACK of message ID 6 Message identifier of the message being ACKed (all 0 s if unknown) RTS flag 1 Request for CoS2 slot(s) Priority (0 to 14) 4 Unassigned if RTS flag is 0 Reservation length 5 Number of slots required (1-20) FEC 17 30% FEC coding CRC 16 ISO flag 8 Binary bits The reservation length field shall be set to 0 if the RTS flag is 0. ACK/CTS message to all aircraft: CTS_ACK_ALL This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 GS local address 7 Message type 6 Binary A/C local address 9 CTS flag 1 Binary 1 indicates CTS RTS message identifier 6 1 to 64 ( to ) Priority (0 to 14) 4 Unassigned if RTS flag is 0 Reserved slot number 9 Binary if no CTS, Reservation length 5 Max value 20 slots, 0 if no CTS Power Control Field for not monitor, range 1-25 ACK flag 1 Binary 0 for NACK, binary 1 for ACK ACK of message type 6 Message type acknowledged (all 0 s if unknown) ACK of message ID 6 ID of message being ACKed (all 0 s if unknown) FEC 0.3 *(16+52n) 30% FEC coding CRC 16 ISO flag 8 Binary n+0.3 *(16+52n) bits Edition: 1.0 Draft Page 94

95 The highlighted box is repeated according to the number of aircraft (n) to which an ACK or CTS is being sent. GS ACK uplink message: GS_ACK This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 GS local address 7 Message type 6 Binary A/C local address 9 Power Control Field for not monitor, range 1-25 ACK flag 1 Reply to CoS2 downlink ACK of message type 6 Message type acknowledged (all 0 s if unknown) ACK of message ID 6 ID of message being ACKed (all 0 s if unknown) FEC 15 30% FEC coding CRC 16 ISO flag 8 Binary bits A NACK to a transmitted message shall trigger a re-send by the mobile station. A mobile station shall ignore an ACK/NACK for an unrecognized message (type or ID). Edition: 1.0 Draft Page 95

96 CoS2 downlink message: DATA_COS2 This message shall be transmitted by a mobile station in CoS2 that has a requirement to send data, or an ACK and data, or and ACK and RTS and data, to the ground station. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 A/C local address 9 Message type 6 Binary Message identifier 6 1 to 64 ( to ) GS local address 7 ACK flag 1 Reply to UP2 uplink ACK of message type 6 Message type acknowledged (all 0 s if unknown) ACK of message ID 6 ID of message being ACKed (all 0 s if unknown) RTS flag 1 Request for CoS2 slot(s) Priority (0 to 14) 4 Unassigned if RTS flag is 0 Reservation length 5 Number of slots required (1-20) Data length (octets) 11 Range 1-2,048 Data Variable length FEC % FEC coding CRC 16 ISO flag 8 Binary bits The ACK fields shall not be included in the message if the ACK flag is set to 0. The RTS fields shall not be included in the message if the RTS flag is set to 0. A/C ACK downlink message: AC_ACK This message shall be transmitted by a mobile station in CoS2 that has a requirement to send an ACK to the ground station. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 A/C local address 9 Message type 6 Binary GS local address 7 ACK flag 1 Reply to UP2/CoS2 uplink ACK of message type 6 Message type acknowledged (all 0 s if unknown) ACK of message ID 6 ID of message being ACKed (all 0 s if unknown) FEC 12 30% FEC coding CRC 16 ISO flag 8 Binary bits A NACK to a transmitted message shall trigger a re-send by the ground station. The ground station shall ignore an ACK/NACK for an unrecognized message ID. Edition: 1.0 Draft Page 96

97 CoS2 random access long data message: DATA_RA_COS2 This message shall be transmitted by a mobile station in CoS2 that has a requirement to send data to the ground station using the random access procedure. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 A/C local address 9 Message type 6 Binary Message identifier 6 1 to 64 ( to ) GS local address 7 Data length (octets) 11 Range 1-2,048. Length 0 is invalid. Data Variable length FEC % FEC coding CRC 16 ISO flag 8 Binary bits Edition: 1.0 Draft Page 97

98 Framing message: GS_TADVANCE This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 GS local address 7 Message type 6 Binary A/C local address 9 T5 Parameter 5 Range 0 to 96 FEC 0.3 *(16+14n) 30% FEC coding CRC 16 ISO flag 8 Binary n+0.3 *(16+14n) bits The A/C local address and corresponding T5 parameter are repeated for n aircraft logged in to the GS cell. GS Power Control message: GS_ACK This message shall be transmitted by a ground station in UP1 or UP2. ISO flag 8 Version number 2 Binary 00 Address length flag 1 Binary 0 GS local address 7 Message type 6 Binary A/C local address 9 Power Control Field for not monitor, range 1-25 FEC 11 30% FEC coding CRC 16 ISO flag 8 Binary bits Edition: 1.0 Draft Page 98

99 ANNEX 3 System operations The following text indicates the possible sequences of events. A3.1 Downlink Each aircraft has an allocated CoS1 slot. A3.1.1 A/C login Note. This represents the communication initiation when an aircraft contacts a new ground station under any circumstances. The aircraft shall use its knowledge of the area to identify the frequency of the ground station whose cell it is about to enter. The aircraft shall listen on this frequency for a minimum of 2 seconds to hear the framing message (GS_FRAME) transmitted by the ground station. Note. This framing message contains the ICAO address of the ground station, the current sizes of sections within the frame and the UTC time and date. When the aircraft has heard the GS_FRAME message, it transmits a CELL_LOGIN message in one of the dedicated LoG2 slots. The CELL_LOGIN message contains the ICAO address of the aircraft. If the GS correctly receives the CELL_LOGIN message, it shall transmit a GS_ALLOC message to the aircraft in UP1 or UP2. Note. The GS_ALLOC message contains the new local 9-bit address for the aircraft, the local 7-bit address of the ground station and the location of the exclusive high-qos CoS1 slot for the aircraft. If the aircraft does not receive the GS_ALLOC message correctly then it shall re-transmit the CELL_LOGIN message in one of the dedicated login slots If the GS does not receive the CELL_LOGIN message correctly, it shall take no action except wait for the aircraft to re-transmit. Edition: 1.0 Draft Page 99

100 A3.1.2 Aircraft has data to send If the aircraft has data to send, the aircraft shall transmit an RTS_COS1 message with RTS flag set to 1. Note. This requests reservation of a specified number of slots in CoS2. If the RTS_COS1 transmission is correctly received by the GS, the GS responds in the combined CTS_ACK_ALL message in UP1 or UP2, setting the ACK flag to 1, indicating which message is being acknowledged, setting the CTS flag to 1 and indicating the allocated slot(s) for the data in CoS2. If the CTS in the CTS_ACK_ALL message is not correctly received by the aircraft then it shall retransmit an RTS for the data in CoS1 or using a DATA_CoS2 message with RTS field set to 1. If the CTS in the CTS_ACK_ALL message is correctly received by the aircraft then it shall transmit the data in the allocated slot(s) in CoS2. If the aircraft does not receive an ACK from the GS for the data transmitted in the next CTS_ACK_ALL message then it shall re-transmit an RTS for the data in its CoS1 slot or using a DATA_CoS2 message with RTS field set to 1 If the DATA_COS1 transmission is not correctly received by the GS, the GS sets the ACK flag to 0 in the CTS_ACK_ALL message in UP1 or UP2 and indicates which message is being NACKed. Upon receipt of a NACK from the GS, the aircraft shall re-transmit an RTS for the data in CoS1 or using a DATA_CoS2 message with RTS field set to 1. A3.1.3 Aircraft has no data to send If the aircraft has no data to send and wishes to maintain the link with the GS, it shall transmit a KEEP_ALIVE message in its CoS1 slot if required to maintain the aircraft s mandatory rate for CoS1 transmissions of at least one transmission per S1 frames. A3.1.4 CoS2 random access If an aircraft has one or more messages in its transmit queue at the start of the CoS2 section of the frame, the aircraft shall identify from its previous reception of GS transmissions all CoS2 slots in the current frame that are free and select a slot or slots from the available free slots for random access according to the random access procedure specified in Section 4.6. If slots are successfully selected according to the random access procedure, the aircraft shall transmit in the selected slots in CoS2 according to the random access procedure using the DATA_RA_COS2 message, and remove the data transmitted from its transmit queue. A GS in receipt of data transmitted by an aircraft by random access in CoS2 shall transmit an ACK in the UP1 or UP2 section of the frame. An aircraft receiving an ACK from the GS for data transmitted by random access shall remove the data transmitted from its transmit queue. An aircraft receiving an NACK from the GS, or which receives no ACK from the GS for data transmitted by random access in the following UP1 or UP2 frame sections, shall maintain the data in its transmit queue for subsequent transmission. A3.1.5 Aircraft-initiated cell exit If the aircraft is about to leave the cell of the GS it shall transmit an AC_CELL_EXIT message in its dedicated CoS1 slot. If the GS correctly receives the CELL_EXIT message it shall transmit a GS_EXIT_ACK message in UP1 or UP2, with the ACK flag set to 1, indicating an ACK slot in CoS2 for the aircraft to confirm. Edition: 1.0 Draft Page 100

101 If the aircraft correctly receives the GS_EXIT_ACK message, the aircraft shall transmit an AC_ACK in the allocated CoS2 slot, with the ACK flag set to 1, the message ID field set to CELL_EXIT and the repeat number field set to 000. If the aircraft does not (correctly) receive the GS_EXIT_ACK message and can still contact the GS, the aircraft shall transmit an AC_ACK in the allocated CoS2 slot, with the ACK flag set to 0, the ACK of message ID field set to CELL_EXIT and the repeat number field set to 000. The aircraft shall then attempt to re-transmit the AC_CELL_EXIT message in its dedicated CoS1 slot in the next frame. If the aircraft correctly receives the GS_EXIT_ACK message and is out-of-range of the GS, the aircraft shall do nothing, and its dedicated CoS1 slot at the old GS will time-out. If the GS receives an AC_ACK message from the aircraft with the ACK flag set to 1, the aircraft's dedicated CoS1 slot shall be de-allocated and the hand-over will be complete. If the GS does not receive an AC_ACK, or receives an AC_ACK with the ACK flag set to 0, the GS shall do nothing. If the aircraft is still in range then the CELL_EXIT shall be retransmitted in the next frame, otherwise the link will time-out. If the GS does not correctly receive the AC_CELL_EXIT message, the GS shall do nothing. If the aircraft is still in range then it shall retransmit the CELL_EXIT in the next frame, otherwise the link will time-out. A3.1.6 GS request for aircraft cell exit with no recommendation If the aircraft is about to leave the cell of the current GS, or the current GS is overloaded, and the GS wishes to trigger a hand-over procedure, the current GS shall transmit a GS_CELL_EXIT message in UP1 or UP2. If the aircraft correctly receives the GS_CELL_EXIT message from the current GS, then the aircraft will determine the frequency of the new GS to contact in the next cell. The aircraft shall attempt to hear the new GS on the expected frequency. If the aircraft receives a framing message from the new GS then it shall commence the cell login procedure (see Section A3.1.1). If the cell login procedure is successful, the aircraft shall transmit an AC_EXIT_ACK message to the current GS in its dedicated CoS1 slot, with the ACK flag set to 1 and the ACK slot number field set to 0. If the current GS receives the AC_EXIT_ACK message correctly, it shall reply with a GS_ACK message with the ACK flag set to 1 and the message ID field set to EXIT_ACK. It shall then de-allocate the aircraft's CoS1 slot and consider the link to be terminated. If the aircraft receives the GS_ACK with the ACK flag set to 1 it shall consider the link to be terminated. If the aircraft receives the GS_ACK with the ACK flag set to 0 it shall re-transmit the AC_EXIT_ACK. If the aircraft does not receive a GS_ACK message it shall still consider the "exit" process to be complete. Note. This is to avoid the situation where a GS_ACK could be transmitted but not received and the aircraft thinks that its old CoS1 slot is still available when in fact it had been reallocated by the GS. If the cell login procedure with the new GS is not successful or the aircraft cannot successfully communicate with the new GS on the new frequency then the aircraft will transmit an AC_EXIT_ACK message to the current GS in its dedicated CoS1 slot, with the ACK flag set to 0 and the ACK slot number field set to 0. Edition: 1.0 Draft Page 101

102 If the current GS receives an AC_EXIT_ACK message with the ACK flag set to 0, or does not (correctly) receive an AC_EXIT_ACK at all, then it shall re-transmit the GS_CELL_EXIT message in the next frame, unless the aircraft transmits an AC_CELL_EXIT message beforehand. If the aircraft does not correctly receive the GS_CELL_EXIT message then it shall transmit an AC_EXIT_ACK message to the current GS in its dedicated CoS1 slot, with the ACK flag set to 0 and the ACK slot number field set to 0. Note. The re-transmission of CELL_EXIT messages by the current GS will only continue while the communications link is still open. If the link times-out then the current GS will consider the aircraft to have left the cell and will de-allocate the aircraft's CoS1 slot. A3.1.7 GS request for aircraft cell exit with recommendation Note. This procedure can only be followed when the aircraft is located within the coverage of several ground stations which are all part of the same cluster. If the aircraft is about to leave the cell of the current GS, or the current GS is overloaded, and the GS wishes to trigger a hand-over procedure, and from data received from the GNI the current GS believes that the aircraft should preferentially hand-over to a specific GS, then the current GS shall transmit a GS_CELL_EXIT message in UP1 or UP2, indicating the GS to which the aircraft should hand-over. If the aircraft correctly receives the CELL_EXIT message from the current GS, and recognizes the suggested GS, then the aircraft shall determine the frequency of the suggested GS to contact. The aircraft will attempt to hear the suggested GS on this frequency. If the aircraft receives a framing message from the suggested GS then it shall commence the cell login procedure (see Section A3.1.1). If the cell login procedure is successful, the aircraft will transmit an EXIT_ACK message to the current GS in its dedicated CoS1 slot, with the ACK flag set to 1 and the ACK slot number field set to 0. If the current GS receives the AC_EXIT_ACK message correctly, it shall reply with a GS_ACK message with the ACK flag set to 1 and the message ID field set to EXIT_ACK. It shall then de-allocate the aircraft's CoS1 slot and consider the link to be terminated. If the aircraft receives the GS_ACK with the ACK flag set to 1 it shall consider the link to be terminated. If the aircraft receives the GS_ACK with the ACK flag set to 0 it shall re-transmit the AC_EXIT_ACK. If the aircraft does not receive a GS_ACK message it will still consider the "exit" process to be complete. Note. This is to avoid the situation where a GS_ACK could be transmitted but not received and the aircraft thinks that its old CoS1 slot is still available when in fact it had been reallocated by the GS. If the aircraft does not recognize the address of the suggested GS, or the cell login procedure with the suggested GS is not successful, then the aircraft shall determine the frequency of another possible GS (determined from location). The aircraft shall attempt to hear the GS on this frequency. If the aircraft receives a framing message from the GS, then it shall commence the normal login procedure for this new GS (as in A3.1.1). If the login procedure is successful then the hand-over shall be considered to be complete. If the login procedure is not successful then the aircraft will attempt to contact another Edition: 1.0 Draft Page 102

103 possible GS. If the cell login procedure with all possible future GS's is not successful then the aircraft shall transmit an AC_EXIT_ACK message in its dedicated CoS1 slot, with the ACK flag set to 0 and the ACK slot number field set to 0. If the GS receives an AC_EXIT_ACK message with the ACK flag set to 0, or does not (correctly) receive an AC_EXIT_ACK at all then it shall re-transmit the GS_CELL_EXIT message in the next frame, unless the aircraft transmits a AC_CELL_EXIT message beforehand. If the aircraft does not correctly receive the GS_CELL_EXIT message then it shall transmit an AC_EXIT_ACK message in its dedicated CoS1 slot, with the ACK flag set to 0 and the ACK slot number field set to 0. If the GS receives an AC_EXIT_ACK message with the ACK flag set to 0, or does not (correctly) receive an AC_EXIT_ACK at all then it shall re-transmit the GS_CELL_EXIT message in the next frame, unless the aircraft transmits a AC_CELL_EXIT message beforehand. Note. The re-transmission of GS_CELL_EXIT messages by the GS will only continue while the communications link is still open. If the link times out then the GS will consider the aircraft to have left the cell and will de-allocate the aircraft's CoS1 slot. A3.2 Uplink Each GS has dedicated uplink sections in the frame. A3.2.1 GS has data to send A Data size is 2,048 octets, if transmitted in UP1 If the data size is 2,048 octets the GS shall transmit a GS_DATA message to the aircraft with the ACK slot number field set to , the repeat flag set to 0 and the repeat number field set to 000. If the UP1 transmission is correctly received by the aircraft, the aircraft shall send an ACK to the ground station, either in an RTS_COS1 message or in a DATA_COS2 message, setting the ACK flag to 1, and indicating which message is being acknowledged. If the UP1 transmission is not correctly received by the aircraft, the aircraft shall send a NACK to the ground station, either in an RTS_COS1 message or in a DATA_COS2 message setting the ACK flag to 0, and indicating which message is being NACKed. If the GS does not correctly receive an acknowledgement for the GS_DATA, or receives a NACK, the GS shall re-transmit the GS_DATA message in UP1 or UP2, with the repeat flag set to 1 and the repeat number incremented from the previous attempt. A Data size is >2,048 octets If the data size is >2,048 octets, the GS will split the data into several GS_DATA messages (each one of which will contain up to 2,048 octets of data). These GS_DATA messages will then be transmitted separately. The method for this is implementation-dependent and is outside the scope of this document. A3.3 Acknowledgement messages An acknowledgement flag shall always be followed by a field indicating the message type. If a station (GS or aircraft) is sending an ACK/NACK message then the message type field Edition: 1.0 Draft Page 103

104 shall be set to the message type which is being acknowledged. If a station (GS or aircraft) is sending an ACK/NACK message then the message identifier field shall be set to the identifier of the message which is being acknowledged. If an ACK or NACK is received by a station (GS or aircraft) and the message type field does not match any of the messages for which an ACK or NACK is expected, then it shall be ignored. The station shall then wait for another ACK/NACK message. If one is not received within a given time 12, this shall be treated as a NACK. If an ACK or NACK is received by a station (GS or aircraft) and the message type field matches one or more message for which an ACK or NACK is expected but the message identifier field does not also match the identifier number of any message or messages for which an ACK or NACK is expected, then it shall be ignored. The station shall then wait for another ACK/NACK message. If one is not received within a given time 13, this shall be treated as a NACK. 12 This requires definition of a timer, which is considered to be outside the scope of this document. 13 This requires definition of a timer, which is considered to be outside the scope of this document. Edition: 1.0 Draft Page 104

105 ANNEX 4 Coding and interleaving This annex provides first results on the coding scheme performance and robustness. Simulations and tests should be performed to validate the performance. This evaluation has been done with the assumptions of a 1366 bits slot length. The performance is even better considering longer slot lengths. A4.1 GMSK and convolutional coding: rate and expected performances We propose here a quick evaluation of the performance for some punctured Convolutional codes with a GMSK. This discussion is based on existing results that can be found in the literature. These are used to derive some performances which could be expected, but this approach requires running complete simulation to assess definitely the performances of the system. Uncoded GMSK is quite similar to MSK (a 0.5 db degradation is observed, in theory, due to ISI introduced by pre-filtering) when a coherent demodulator is applied (see [10] for a complete discussion). This curve below has been obtained by shifting the MSK (with coherent demodulation) performance by 0.5 db. Edition: 1.0 Draft Page 105

106 Figure 30: GMSK theoretical performance The convolutional code used on the next figure is the well known (LC = 3, (5,7), d free = 5), and is represented below (LC for Constraint Length). The simulated performances are for a BPSK over AWGN channel. The figure above provides an idea of the BER which can be expected at the output of the Viterbi decoder, for a given BER at the input. One can check that the ratio is quite linear (with a log-log scale) for BER at the input smaller than These well-known results can be found in [11], chap. 8. Edition: 1.0 Draft Page 106

107 Figure 31: Convolutional code performance G 2 = 5 (8) c 2 b k b k-1 b k-2 D D G 1 = 7 (8) Edition: 1.0 Draft Page 107 c 1

108 Figure 32: Convolutional code (5,7), constraint length 3 For the code LC = 7, (171,133), the free distance d free = 5 corresponds to a rate ¾. As performance of a code depends mainly on its free distance when the SNR increases, the two codes can be considered equivalent in terms of performances, though mostly if SNR is high enough. This code is represented below. input bit G 2 = 133 (8) b k b k-1 b k-2 b k-3 b k-4 b k-5 b k-6 D D D D D D output 2 output 1 Figure 33: Convolutional code (133,171), constraint length 7 G 1 = 171 (8) The table below shows the relation between coding rate and corresponding free distance.for punctured convolutional codes, based on (LC = 7, 171,133). R c ½ 2/3 ¾ 5/6 7/8 d free Table 2: Free distance for some code rate Based on these considerations, we can derive the BER at the input of the coder from Figure 30: GMSK theoretical performance, and then derive the BER after decoding from Figure 31: Convolutional code performance. Example of application to L-DACS2: C/N = Eb/N log (se), where se is the spectral efficiency. R C = ½, se = 270/200 If C/N = 10 db then Eb/N0 = 8.7 db This leads to a BER (undecoded) equal to , and the BER (decoded, inner code) is then , which leads to less than 10-7 if either RS(31,23,5) or RS(15,11,4) are used. A4.2 Considerations regarding practical C/N The following discussion is based on information derived from the link budgets which have been computed for L-DACS2. Receiver threshold: - 96 dbm N = FkTB, with F = 10 db, T =288.4 K, B = 200 khz, N = -108 dbm At receiving threshold, C/N = 12 db. This is 2 db more than used to derive performances in the previous paragraph, and correspond to a significant engineering margin. Though it does not represent a proof of the performances, the fact that a margin is still available strengthens this approach. Edition: 1.0 Draft Page 108

109 A4.3 Accounting for burst interference: interleaving and RS coding In this discussion regarding interleaving and outer coding, we assume a punctured convolutional code (LC = 7, 133,171), with a rate ¾ or 5/6, as inner code. The objective is to evaluate the right parameters for coding and interleaving, in the presence of strong bursts of errors, due to a very near (co-located) interfering transmitter. To ensure independence between two different communications, we limit the depth of the interleaver to a slot length (1366 coded bits). RS notation: RS (N, N-2T,k), where N = 2^k 1, and T is the number of corrected errors. Burst length: Symbol duration 3,69231E-06 Duration (ms) Duration (bits) Burst, type Burst, type The slot length is 1366 bits, equivalent to 1024 bits before convolutional encoding with a rate ¾., so one slot is at most 7 codeword long, if RS(31,X,5) is used. For a burst of 114 bits, with bloc interleaving, it is likely that about 16 consecutive bits will be strongly errored after de-interleaving. This leads to 4 errors in a codeword. RS(31,23,5) should thus be considered rather than RS(31,29,5) or RS(31,27,5). Another option is to use RS(15,11,4), for which a code word is 60 bits long, leading to about 18 codewords in a slot. After de-interleaving, 6 consecutive bits at most are errored, and correspond to 2 to 3 errors in a codeword. BER_in RS(31,29,5) RS(31,27,5) RS(31,25,5) RS(31,23,5) RS(15,13,4) RS(15,11,4) 5,0E-03 2,2E-03 4,8E-04 7,9E-05 1,0E-05 6,0E-04 5,9E-05 1,0E-03 2,5E-05 1,2E-06 3,9E-08 1,0E-09 5,8E-06 1,2E-07 2,0E-04 2,0E-07 2,1E-09 1,4E-11 7,2E-14 4,8E-08 1,9E-10 1,0E-04 2,8E-08 1,3E-10 4,4E-13 1,1E-15 6,0E-09 1,2E-11 1,0E-05 2,8E-11 1,3E-14 4,4E-18 1,1E-21 6,1E-12 1,2E-15 Code word length 155 bits 155 bits 155 bits 155 bits 60 bits 60 bits Rate Table 3: Performances of various RS codes The results in the table above are for uncorrelated errors, and cannot be applied straightforwardly to burst of errors. The presence of an ideal interleaver would solve this problem, but cannot be implemented in practice. Anyway we may consider these results as providing an idea of the performances which may be reached, though dedicated simulations, modelling both the interfering burst signal and the propagation channel are mandatory to derive accurate results. Edition: 1.0 Draft Page 109

110 A4.4 Equalization L-DACS2 System Definition proposal: Deliverable D2 Filtering the signal (with the Gaussian filter) introduces inter symbol interference (ISI), even over a single path channel. This is the reason why GMSK is less efficient than MSK. In the GSM system, equalization is used to fight ISI, but it is true that ISI in GSM is due to both pre-filtering and multipaths propagation. The influence of equalization should be considered for L-DACS2 as well, probably through channel estimation and MLSE equalizer (Maximum Likelihood Sequence Estimate, Viterbi algorithm using equivalent discrete channel estimation for weighting). This could improve performance at the cost of introducing a training sequence (CAZAC sequence is used in GSM, but PRBS can also be used). Edition: 1.0 Draft Page 110

111 ANNEX 5 Impact of intra-system interference A5.1 Intra system interference robustness An evaluation of the co-channel impact has been conducted in order to have statistic on the impact of deployment of the system over the link budget. The link budget presented with co-channel interference figure should be considered with caution. It is related to the channel deployment of the system. Currently the attenuation of the interfered signal coming from a cell using the same frequency is 9 db based on a 12 frequency reused pattern and the performance of the GMSK modulation. The degradation caused by a man-made signal, with the same characteristics (modulation, frequency, bandwidth, etc.) on the signal of interest, is smaller than the degradation caused by additive white Gaussian noise but for this exercise it was assimilated to additive white Gaussian noise. Additional test on the prototype should be performed in order to validate the current attenuation of a co-channel signal and the accurate distance needed to use frequency reused channel. First analysis is conducted hereafter in order to give first input for the frequency plan of the L-DACS2 system. Nevertheless those inputs need to be validated during the prototype phase. The overall impact of one channel interfering in the same frequency band could deeply reduce the net margin considering the operational case where the interfering aircraft is close to the station and still transmits with the maximum power. Power control should decrease this impact for those cases. Edition: 1.0 Draft Page 111

112 UL DL TX Parameters Unit ENR ENR ENR TMA APT ENR ENR ENR TMA APT Governing Equation Notes L-DACS2 TX ouput Power dbm 55, , , , , , , , , ,9897 a Tx_Pout - (UL: 350W) - (DL: 50W) Maximum TX Antenna Gain dbi b TX_AntGain - (UL: DME antenna reference) - (DL: Antenna Gain = 3dB - polarization loss = 3dB) Tx Cable loss db 2,5 2,5 2,5 2,5 2, c TX_CableLoss - (UL: 50* *0.25) TX EiRP dbm 60, , , , , , , , , ,9897 d =a + b - c TX EiRP = TX_Pout + TX_AntGain - TX_CableLoss Propagation Parameters Transmit upper Frequency MHz e Tx-Rx Distance Nm f Path Loss db 143,62 139,18 133,16 129,64 117,60 143,62 139,18 133,16 129,64 g = 37,8 + 20log(f) + 117,60 20log(e) Free Space model (using nm unit) Miscellaneous Margins Interference Margin db h InterfMargin(TBD) Implementation Margin db i ImpMargin Safety Margin db j SafetyMargin(TBD) Banking Loss Margin db k Banking(TBD) RX Parameters Maximum RX Antenna Gain dbi l RX_AntGain - (UL: Antenna Gain = 3dB - polarization loss = 3dB) - (DL: DME antenna reference) Rx Cable loss (incl. Duplexer) db ,5 2,5 2,5 2,5 2,5 m RX_CableLoss - (UL: 50* *0.25) L-DACS2 RX receive signal dbm -82,68-78,24-72,22-68,70-56,66-91,63-87,19-81,17-77,65-65,61 n = d - g + l - m RxPower = TX_EiRP - PathLoss + Rx_AntGain - Rx_CableLoss Thermal Noise Density@290K dbm/hz o 10log(kT ) Bandwidth Hz p BW Thermal Noise Power dbm -120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99-120,99 q = o + 10log(p) 10log(k.T) +10log(BW ) Receiver Noise Figure db r Rx_NF Composite Noise Figure db ,5 9,5 9,5 9,5 9,5 z Composite noise including Rx noise and cable loss Total Rx Noise Power dbm -107,99-107,99-107,99-107,99-107,99-111,49-111,49-111,49-111,49-111,49 s = q + z + i + [h + j + k] N = Rx_NF + 10log(k.T.BW) BER=10-3 db t Eb/No L-DACS2 bit rate bps u R Required C/N db 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 v = t + 10log(u/p) C/N = Eb/No + 10log(R/BW) L-DACS2 Rx Sensitivity dbm -96,67-96,67-96,67-96,67-96,67-100,17-100,17-100,17-100,17-100,17 w = v + s Cmin = C/N + N L-DACS2 C/N available db 25,31 29,75 35,77 39,29 51,33 19,86 24,30 30,32 33,84 45,88 n-s L-DACS2 net db 8,00 12,43 18,45 14,97 27,02 2,54 6,98 13,00 9,52 21,56 n-s-v Considering ImpMargin and [other margins above] RX Parameters with interference Duty Cycle % 3,00 3,00 3,00 3,00 3,00 3,00 3,00 3,00 3,00 C/I: Frequency reuse interference assumption db z Interference I due to frequency reuse dbm -106,91-102,47-96,45-92,93-80,89-115,86-111,42-105,40-101,88-89,84 x 10log(10^((n-z)/10)*@/100) Total Rx Noise Power without N = Rx_NF + 10log(k.T.BW) + ImpMargin + [other margins interference dbm -101,99-101,99-101,99-94,99-94,99-105,49-105,49-105,49-98,49-98,49 s = q + z + i + [h + j + k] above] Noise increase due to interference dbm 1,21 2,78 6,61 4,16 14,27 0,38 0,99 3,06 1,64 9,21 10log(10^(s/10)+10^(x/10))-s Total Rx Noise Power dbm -100,78-99,21-95,38-90,83-80,72-105,11-104,50-102,43-96,85-89,28 y s+noise increase BER=10-3 db t Eb/No L-DACS2 bit rate bps u R Required C/N db 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 11,32 v = t + 10log(u/p) C/N = Eb/No + 10log(R/BW) L-DACS2 Rx Sensitivity dbm -89,46-87,90-84,06-79,51-69,40-93,79-93,19-91,12-85,53-77,97 w = v + s Cmin = C/N + N L-DACS2 C/N available db 18,10 20,97 23,16 22,13 24,06 13,48 17,31 21,26 19,20 23,67 n-y L-DACS2 net db 6,78 9,66 11,84 10,81 12,75 2,16 5,99 9,95 7,89 12,36 n-y-v Table 29 Link budget including interference contributions Edition: 1.0 Draft Page 112

113 A5.2 Adjacent channel frequency guard time Based on the adjacent band reference interference ration presented in section and using on operational scenario, an estimation of the frequency band needed for adjacent channel was conducted. The operational scenario considered is two airplanes logged in two adjacent cells and located in the border of the cell (at the maximum range of the cells). One aircraft receives a transmission from the ground (left column) and from the other airplane (right column) The reference interference ratio C/I a (I a for Interference adjacent) for the ground and airborne receiving function are listed in the following table. for adjacent (200 khz) interference for adjacent (400 khz) interference for adjacent (600 khz) interference C/Ia -9 db -41 db -49 db Two evaluation were conducted Using 200 khz adjacent frequency (left table) Using 400 khz adjacent frequency (right table) In order to ease the frequency plan of the L-DACS2 system only multiple of 200 KHz (the bandwidth of one channel) are studied. Table 30 Adjacent channel frequency guard time In the 200 KHz scenario case, the distance between the two mobiles has to be 10 Nm in order to respect the interference ratio. Considering the 400 KHz scenario, the receiver could still demodulate the ground signal even if the distance between the two mobiles is 0.23 Nm. This distance is far less than the required separation between two mobiles. Therefore the frequency band for adjacent channel has to be fixed at 400 KHz at minimum. A5.3 Co-channel spacing guard band For Co-channel frequency band, the operational scenario takes into account two mobiles Edition: 1.0 Draft Page 113

114 operating in the same channel but not in the same cell. One mobile is receiving a transmission form the ground while the other one is transmitting. The objective is to get the minimum distance between the two mobiles that leads to the demodulation of the signal transmitted by the ground to one mobile while the other one transmits. The figure below demonstrates that 80 Nm is required in order to respect the 9 db of C/I c (I c for interference co-channel) Table 31 Co-channel spacing guard band A5.4 Multiple channels operating in one cell A5.4.1 Frequency Band The same kind of evaluation has been processed for co-channel cohabitation. The objective of this evaluation is to fix the frequency guard time between two channels assigned in the same cell. The same kind of operational scenario that above has been used for co-channel frequency band considering than the two mobiles could be even closer ft has been selected as the minimum distance between two mobiles. The figure below shows the level of received signal in this scenario. Edition: 1.0 Draft Page 114

115 Table 32 Multiple channels operating in one cell scenario The C/I ratio obtained in this scenario is db. Therefore according the table presented in section , a 600 KHz frequency band at minimum has to be set between two operational channel operating in the same cell. A5.4.2 Spacing separation between stations Although the use of adjacent channels introduces the probability of a finite amount of interference, undesired cross-talk effects across the two or more channels operating in the cell can be minimised through the use of sitting contingencies. On the ground side, appreciable isolation is possible by means of the directivity properties of sectored antennas and the use of frequency planning rules (for example, a minimum of two channels separation within the same ground site). There is no requirement for any specific treatment of a two (or more) channel system on the aircraft, since this aircraft will simply tune to the single channel to which it has been assigned by the ground system. Note however that the use of adjacent channels in the same cell is not a nominal case, and that such case is a planning constraint rather than design constraint. The adjacent protection specifications allow for the signal in the adjacent channel to be 9 db greater than the desired signal for acceptable BER. The constraints imposed by the operation of two adjacent channels are illustrated in the following figure. In this extreme scenario, the ground station (operating across two channels) has active links with an aircraft in the near vicinity on one channel (for example, on the airport surface), and with a second aircraft in the far field on the second channel (for example, at the edge of the TMA or En route sector). Edition: 1.0 Draft Page 115