ACP WGC7/WP AERONAUTICAL COMMUNICATIONS PANEL (ACP) Working Group-C - 7 th meeting Montreal, Canada 19-23 April 2004 Agenda item : Concept of Self-synchronized Automatic Dependent Surveillance using Satellite INFORMATION PAPER Presented by Yasuto Sumiya (Prepared by Yasuto Sumiya, Akira Ishide) Electronic Navigation Research Institute (ENRI), Japan SUMMARY Satellite-based Automatic Dependent Surveillance (ADS) using a polling protocol can reduce a transmission delay greatly compared with the conventional ADS using the satellite data communication that meets ICAO Aeronautical Mobile Satellite Services (AMSS) standards. Although the polling protocol shows excellent performances, it still has some shortcomings. In the polling ADS, an ADS report request must be transmitted for every ADS report, and the guard time between successive ADS reports must be larger than the maximum difference of round-trip times between a satellite and an AES. These cause some deterioration in transmission efficiency of the channels. This report describes self-synchronized ADS using a satellite. It improves the transmission efficiency over the channels of both ADS report request and ADS report. Since any modulation and coding schemes can be used with this protocol, it can also be applied to future aeronautical satellite communication systems.
1. Introduction The Automatic Dependent Surveillance (ADS) using a satellite has recently been introduced to the oceanic air traffic control for surveillance. The ADS is a surveillance system by which aircraft positions derived from on-board navigation system, automatically and periodically, are transmitted through a data link to an air traffic control center. Air traffic controllers can monitor the positions of aircraft flying over the ocean accurately in real-time using ADS. The ADS report transfer through a data link that meets present ICAO AMSS SARPs suffers a substantial delay and the delay may further increase due to the possible traffic congestion [1]. We have proposed a simple polling protocol for ADS that reduces a transmission delay substantially. By using the protocol, the delay can be reduced to between one and two seconds [2]. Although the protocol gives excellent performances, it still has some shortcomings. The followings are the shortcomings of the protocol in relation to the efficiency of channel usage. a) The channels for ADS report request and ADS report must be paired with one-to-one correspondence. b) The channel throughput decreases as the channel rate increases due to guard times. This report describes self-synchronized ADS using a satellite that promote more efficient use of channels. 2. Polling ADS 2.1. Concept Figure 1 shows the concept of ADS using a polling protocol through a satellite (hereinafter referred to as polling ADS ). In this concept, an ATC center sends an ADS report request to each AES in the coverage of a spot beam sequentially. Then, each AES sends back an ADS report to the ground immediately in response to the request. The use of spot beams reduces the required gain of AES antenna. The interval of ADS reports for each AES is 10 seconds, the same value of the Second Surveillance Radar (SSR). When a Cyclic Redundancy Code (CRC) indicates errors in the received ADS report, it is discarded. The polling protocol minimizes the transmission delay and prevents its increase, and keeps the receipt interval almost constantly. 2.2. Transmission Sequence Figure 2 shows the data transmission sequence for the polling ADS. The guard time is required to avoid collisions of ADS reports sent from different aircraft on the same channel. It is calculated from the maximum difference of round-trip times between the satellite and all AES. 2.3. Data Format Figure 3 shows an example of an ADS report format. It is the format specified for R-channel Signal Unit (SU) in the AMSS SARPs. We assume this format in this paper for convenience because it can accommodate the required ADS report items (10 octet) without modification. Since the format includes unnecessary information, it should be modified appropriately later. To minimize the data size, the ADS report only includes a basic ADS block that consists of latitude, longitude, altitude, time and FOM (Figure Of Merit). The FOM is defined as an indicator of positioning accuracy of on board navigation system. The CRC is used to detect errors in a received SU. If any errors are found in the received SU, it is discarded. Figure 4 shows an example of a burst signal. The modulation and coding schemes specified for R-channel in the AMSS SARPs are assumed. We assume them for convenience, because we can 1
estimate the transmission performances such as a bit error rate, delay, etc. accurately from the results of experiments and simulations we conducted. If more advanced modulation and coding techniques are applied, the performances can be further improved. The assumption for a format, modulation and coding is sufficient for our purposes to show the usefulness of the new ADS protocol. 2.4. Shortcomings in Polling ADS Figure 5 shows the transmission flow for the polling ADS. The burst length is 0.1269 seconds for 4.8 kbps, and 0.0846 for 10.5 kbps. The guard time is assumed 0.04 seconds. This figure implies possible shortcomings of the ADS. a) The channels for ADS report requests and ADS reports must be paired with one-to-one correspondence. b) The effect of guard time on the efficiency of channel usage increases with the channel rate. 3. Self-synchronized ADS We propose a new ADS protocol in this paper. This protocol supports more efficient use of channels compared with the polling protocol. 3.1. Concept Figure 6 shows the concept of self-synchronized ADS using a satellite. We assume to use spot beams so that small airplanes with a low gain antenna can communicate with a GES for a channel rate of 4.8 kbps or more. The content, the data format and the burst format of ADS reports are assumed to be the same as those of the polling ADS. Each AES synchronizes to UTC by using an on-board GPS receiver in the self-synchronized ADS, whereas each AES is synchronized to the frame of a received P channel signal in the polling ADS. Figure 7 shows the slot structure for ADS report for a channel rate of 4.8 kbps. The frame is 10 second long and each frame begins at the start of every 10 seconds in UTC. Each frame is divided into 66 slots for a channel rate of 4.8 kbps. The slot length is determined from the burst length (0.1269 s) and the guard time (0.02 s). When an aircraft enters the coverage of a satellite spot beam, a vacant slot is allocated for the transmission of an ADS report out of 66 slots. The AES sends ADS reports to the ground every 10 seconds in the allocated slot. For a channel rate of 10.5 kbps, the slot length comes to 0.11 s and the frame contains 90 slots. Figure 8 shows the flow of ADS report transmission. When AES 1 receives the ADS report request addressed to AES 1, the AES transmits ADS reports on an allocated slot thereafter. Then, AES 2 receives the ADS report request addressed to AES 2, and the AES starts ADS report on an earliest slot assigned, and repeats them on successive assigned slots thereafter. Then, AES 3 starts ADS report request and ADS report in the same manner. In this way, each AES that enters the coverage joins one after another. Since the ADS report request or the allocation of a slot occur only once, the channel for ADS report requests is shared by multiple users of different channels for ADS report. Since the transmission timing of ADS report at each AES is synchronized to the same time reference (UTC), the guard time can be reduced to half compared with that of the polling ADS. 3.2. Link Budget In this section, we study the link budget for the case of Multipurpose Transport Satellite (MTSAT) as an example. The satellite G/T at L band is -9 dbk for a global beam and -2 dbk for a spot 2
beam. This implies that the C/N 0 for the link from an AES to a satellite increases 7 db using a spot beam instead of a global beam. Since 1.2 kbit/s data can be transferred from an AES to a GES using a low gain AES antenna in the coverage of a global beam, it is conceivable that the transfer of 6 kbit/s data (5 times of 1.2 kbit/s) may be possible in the coverage of a spot beam. Table 1 lists an example of the link budget for 4.8 kbit/s data transmission. The parameters are determined on a basis of the link budget for MTSAT. The required C/N 0 is the value for 4.8 kbit/s AQPSK quoted from the table of SDM (Table 8.1.1.2). The BER of less than 10-5 is obtained at the C/N 0 of more than the required C/N 0. This link budget shows it is feasible to transfer 4.8 kbit/s data for an AES with a low gain antenna using a spot-beam antenna on the satellite. Figure 9 shows the BER characteristics for 4.8 kbit/s AQPSK obtained from computer simulations. In the simulations, the rician channel is assumed with a fading bandwidth of 60 Hz and with C/M of 10 or 12 db. For a forward channel, the required C/N 0 is larger than that of the simulation result. However, for a return channel, the required C/N 0 is a little less than that of the simulation result. 3.3. Performance Analysis 3.3.1. Transmission Delay In this section, we estimate possible transmission delays. Figure 10 shows the time chart for ADS report on 600 bit/s R channel. The time from CMU processing (Communication channel, The times for GPS processing, data Management Unit) on the AES are obtained from the measurement using the experimental AES equipment. The time from SDU processing on the AES to the GES processing is obtained by using the calculated times of data transfer and propagation and the actual processing time for modulation and demodulation for the experimental equipment. The resultant value of transmission delay comes to about 2.9 seconds for a channel rate of 600 bit/s. We can estimate the transmission delay for other channel rate. Table 2 lists the estimated transmission delay for various channel rates. This table shows that the transmission delay for the channel rate of 4.8 kbit/s is about 1.3 second. The signal format of the self-synchronized ADS is the same as that of the polling ADS and the guard time of the self-synchronized ADS is shorter than that of the polling ADS. The value in the table is almost the same as the value in the case of polling ADS [2]. It should be noted that the values in the table do not include the transmission delay for terrestrial networks. If the transmission delay for terrestrialnetworks is assumed 0.5 seconds (This value was typical when the DDX-P, a typical public packet data network in Japan, was used in the satellite datalink experiment we conducted in 1992-1994.), the total transmission delay will be about 2 seconds. 3.3.2. Throughput and Capacity Figure 11 shows the throughput and the number of AES per channel. Throughput is defined as the percentage in time that burst signals occupy the link in this paper. The polling ADS can deal with a maximum of 59 AES (a throughput of about 75%) for a channel rate of 4.8 kbps, and a maximum of 80 AES (a throughput of about 68%) for a channel rate of 10.5 kbps per channel. The self-synchronized ADS can deal with a maximum of 66 AES (a throughput of about 84%) for a channel rate of 4.8 kbps, and a maximum of 90 AES (a throughput of about 76%) for a channel rate of 10.5 kbps per channel. Thus, the throughput and the number of AES for the self-synchronized ADS are more than those for the polling ADS. 3
3.3.3. ADS Report Loss Rate We define the ADS report loss rate as the percentage of ADS reports that have not reached to the GES out of all transmitted ADS reports. In the polling ADS, there are two cases where an ADS report does not arrive at the GES. One occurs when an ADS report request is transmitted but is not received at the AES, and the other when an ADS report is lost and is not received at the GES. In the former case, the probability of ADS report requests ( M bits) being lost on the link from the GES to the AES P is given by M [ 1 ( 1 P ) ] + ( P ) PRQ = PfUW fb 1 fuw where PfUW is a probability of unique word detection and P fb a bit error rate. In the same manner, the probability of ADS reports ( N bits) being lost on the link from the AES to the GES P REP is given by N PREP = PrUW [ 1 ( 1 Prb ) ] + ( 1 PrUW ) where PrUW is a probability of unique word detection and P rb a bit error rate. Thus, we find the ADS report loss rate P ADS for the polling ADS by P ADS = PRQ + ( 1 PRQ ) PREP. The ADS report loss rate P ADS for the self -synchronized ADS is given by PADS P REP. Because the ADS report request is transmitted only once first in the case of the self-synchronized ADS. Figure 12 shows the ADS report loss rate calculated by the equations above for the polling ADS and for the self-synchronized ADS. The ADS report loss rate is 3x10-3 for the polling ADS and 1.5x10-3 for the self-synchronized ADS assuming the bit error rate for forward and return links is 10-5. This value implies that one report is lost in about 1 hour for the polling ADS and in about 2 hours for the self-synchronized ADS. 4. Conclusion In this paper, we have described the self-synchronized ADS that supports more efficient use of channels compared with the polling ADS. The following merits of the self- synchronized ADS were confirmed over the polling protocol. a) The required number of channels for ADS report request transmissions decreases. b) The channel throughput increases. c) The ADS report loss rate decreases. In this paper, we assumed the modulation and coding schemes specified in the AMSS SARPs as an example. It should be noted that the polling ADS and the self-synchronized ADS can be used with any modulation and coding techniques. References [1] A.Ishide, Update on Comparison of Transmission Delays for Experiments and Simulation (II), 16th Meeting of AMCP WG-A, WP-663, January 2000. [2] A.Ishide, A New Concept of Satellite-based Automatic Dependent Surveillance, Fifth Meeting of AMCP WG-C, WP-18, October 2002. [3] Y.Sumiya, A.Ishide, Concept of Self-synchronized ADS using Satellite, 3rd Meeting of NexSAT, WP-3, October 2003. RQ 4
ADS Report Request N 2 1 1 2 N ADS Report Satellite AES N AES 2 AES 1 ATC Center GES Spot Beams Figure 1. Concept of Polling-type ADS Figure 2. Transmission Sequence of polling ADS Figure 3. ADS Report Format Figure 5. Transmission Flow in polling ADS Air Traffic Center Figure 4. Burst Signal Format GES ADS Report Request N 2 1 1 2 N Satellite ADS Report AES N AES 2 AES 1 Spot Beams UTC GPS Figure 6. Concept of self-synchronized ADS Figure 7. Slot Structure for ADS Report 5
Figure 8. Slot Allocation in ADS Report Transmission Table 1. Example of Link Budget Forward Link Return Link GES e.i.r.p. (dbw) 62.8 AES e.i.r.p. (dbw) 10.5 Propagation Loss (db) 206.6 Propagation Loss (db) 189 Rain Fade (db) 11.8 Satellite G/T (dbk) -1 Satellite G/T (dbk) -2 Up-link C/N 0 (dbhz) 71.9 Up-link C/N 0 (dbhz) 48.1 Satellite e.i.r.p. (dbw) 30.3 Satellite e.i.r.p. (dbw) -1.5 Propagation Loss (db) 188.5 Propagation Loss (db) 205.2 Rain Fade and Additive noise (db) 12 AES G/T (dbk) -26 GES G/T (dbk) 39 Down-link C/N 0 (dbhz) 44.4 Down-link C/N 0 (dbhz) 48.9 Total C/N 0 (dbhz) 44.4 Total C/N 0 (dbhz) 45.5 Required C/N 0 (dbhz) 42.2 Required C/N 0 (dbhz) 42.3 Frequency Band: Ku band (GES-Satellite), L band (Satellite-AES) Required C/N 0 : 4.8kbit/s AQPSK (SDM) 6
10 0 10-1 AWGN(Theory) AWGN C/M=12dB,P-ch C/M=10dB,P-ch C/M=12dB,R-ch C/M=10dB,R-ch Throughput (%) 100 80 60 40 20 Polling 4.8kbps Polling 10.5kbps Self-synchronized 4.8kbps Self-synchronized 10.5kbps Bit Error Rate 10-2 10-3 0 0 20 40 60 80 100 Number of AES Figure 11. Throughput and Number of AES per Channel 10-4 10-5 34 36 38 40 42 44 46 48 C/N 0 (dbhz) 10 0 10-1 10-2 10-3 Figure 9. BER Characteristics obtained from Simulations P ADS 10-2 10-3 10-4 10-5 P fb =10-6 Polling Self-synchronized 10-4 10-6 10-5 10-4 10-3 10-2 10-1 10 0 P rb Figure 12. ADS Report Loss Rate Figure 10. Time Chart for ADS Report on R channel Table 2. Estimated Transmission Delay Channel Rate (kbit/s) Estimated Transmission Delay (s) 0.6 2.9 1.2 1.9 2.4 1.4 4.8 1.3 10.5 1.1 7