FANTASTIC-5G: Novel, flexible air interface for enabling efficient multiservice coexistence for 5G below 6GHz

Transcription

1 FANTASTIC-5G: Novel, flexible air interface for enabling efficient multiservice coexistence for 5G below 6GHz Frank Schaich with support from the whole consortium January

2 Agenda Introduction - Main ambitions of 5G and key challenge - Why a multi-service air interface instead of dedicated solutions? - Use case selection The Project Some exemplary technologies - Waveform design - Enablers for massive access - Flexible frame design and radio resource management Extension of the FANTASTIC-5G air interface to specialized scenarios and extreme cases 2

3 Introduction Main ambitions of 5G and key challenge 1. Finding a response to the strong growth of requested data rates both per connection and overall (evolutionary effect) 2. Enhancing the business model of operators by widening the pool of services (revolutionary target) Key challenge: extremely high degree of heterogeneity in terms of - Offered services (MBB, MMC, MCC, BMS, V2X) and the respective requirements - Supported device classes (from low-end sensor to high-end tablet) - Deployment types (macro layer, small cells) - Environments (low-density areas to ultra-dense urban) - Mobility levels (from static to high-speed transport) FANTASTIC-5G targets to develop a flexible and scalable multi-service air interface (ambition 2) with ubiquitous coverage and high capacity where and when needed (ambition1) 3

4 Introduction Why a multi-service air interface instead of dedicated solutions? Less carriers to be active in parallel More efficient use of the active carrier(s) use case multiplexing, resource sharing Economy of scale shared functions Forward compatibility assured future use cases might deviate from the settings of specialised air interfaces. Specific devices (e.g. smart phones, cars) may comprise different use cases concurrently benefiting from a single harmonized air interface (e.g. embb for streaming, MMC for instant messaging, MCC for mobile gaming) Increased flexibility It is easier to run a multi-service air interface in a specialized configuration than to assure broad applicability with specialized solutions (4G and the need of 5G as example) 4

5 Introduction Use case selection We base our investigations on the following set of use cases (following NGMN recommendations): - 50 Mbps everywhere - High speed train - Sensor networks - Tactile Internet - Automatic traffic control/driving - Broadcast like services: Local, Regional, National - Dense Urban Society (below 6 GHz) The last item is a modified version taking the project considerations into account (to stay below 6 GHz) Our aim has been to have a representative set of 5G use cases, while keeping the number (and thus the overhead) at a reasonable level. 5

6 The Project Flexible Air interface for Scalable service delivery within wireless Communication networks of the 5th Generation Start: July Duration: 2 years Part of the overall 5G PPP framework The participation of multiple vendors (Nokia, Alcatel-Lucent, Huawei), equipment manufacturers (Intel, Samsung, Sequans) and operators (Orange, Telecom Italia) highlights the ambition and capability of having an impact to standardization 6

7 Some exemplary technologies In the following we provide some details on specific components being investigated in the project - Waveform design - Enablers for massive access - flexible frame design and radio resource management 7

8 Waveform design (candidates under study) OOBE: Out of band emissions DoF: degree of freedom SE: spectral efficiency WF: waveform CP: cyclic prefix WF Features Pros Cons Targeted service UF-OFDM / F-OFDM ZT-OFDM FC-OFDM OFDM compatible, Low OOBE for sub-band Good coexist. capabilities thanks to low OOBE Multiplexing of diff. WFs in the transmission band sub-band wise configurability, async. FDMA access support Coexist. with OFDM, zero-tail adjustable Coexistence of diff. WFs in the same band slightly more prone to delay-spread channels Overhead ~ zero tail Constraints on filter length Multi-service MTC All - each service may use its own WF Filtering per sub-band FBMC-OQAM min. OOBE (steepest decay), max. SE, real-field orthogonality sub-band wise configurability, robustness to t/f distortions, full async. access support Not fully OFDM comp., Long filter tails incr. delay for short bursts Multi-service FS-FBMC FBMC-QAM FBMC-OQAM with large FFT for mod. /demod., simple parametrization FBMC with QAM support OFDM compatible Like FBMC-OQAM + supports large delay spreads & enlarged support of time async. sub-band wise configurability, async. FDMA access support Like FBMC-OQAM t/f localization is compromised Multi-service Multi-service Filtering per subcarrier P-OFDM pulse shaping as add. DoF OFDM compatible, low OOBE, CP-like overhead sub-band wise configurability, robustness to t/f distortions, full async. access support complexity slightly higher than OFDM Multi-service 8

9 Enablers for massive access A key characteristic of massive machine access is sporadic traffic from a very high number of sources. Scheduled access procedures (4G) are not very suitable for this (overhead, energy consumption) Simplified but robust access mechanisms need to be developed. Two options are under investigation: - Wake-up and transmit (1-stage procedures) - Wake-up, raise hand and transmit (2-stage procedures) Two aspects requiring deeper investigations: - Protocol design - User identification 9

10 Enablers for massive access protocol design One-stage vs. two stage protocols UE UE enodeb Preamble Data ACK Novel preamble design with potentially more efficient usage of resources enodeb Service request Significant reduction of overhead compared to LTE Info regarding resource assignment Data ACK Faster if successful Significantly less DL feedback High collision probability reduces throughput Coexistence with scheduled traffic difficult to handle Envisaged solution for very small packets and low traffic load Additional delay Depending on configuration certain amount of DL feedback required Reduced collision probability through service request over-provisioning increases throughput Envisaged solution for bigger packets and higher traffic load Both variants require a pool of identifiers (preamble, service request) 10

11 Enablers for massive access protocol design One-stage vs. two stage protocols 1-step protocol is an attractive solution in low-mid load situation Protocols with higher flexibility like the 2-stage protocols are better suited in high load situations. They : a) allow a higher arrival rate and b) achieve a higher cell capacity Both variants benefit from multi-user detection (work in progress) 1-stage protocol is able to upload packets quicker than 2-stage protocols as long the arrival rate is below threshold 2-stage pooled protocol is able to upload packets much quicker than the 2-stage tagged approach 11

12 Enablers for massive access user identification Starting point: Preamble design of 4G (PRACH) - Issue: very small pool of preambles available (preambles designed for very big cells up to 100 km) - Options under investigation: - adapt LTE PRACH preambles using more cyclic shifts (at the cost of higher missed detection rates and false alarms) - apply different root sequences at the cost of mutual interference (between different roots), - m-sequences at the cost of higher PAPR, - CDMA codes at the cost of requiring high synch accuracy. - Alternative approach: shape arrival distribution - slot access principle (map access opportunities to device ID i.e. a given device is only allowed to transmit in predefined sub-frames) at the cost of increased latency - Performance boost (future work): - MUD for collision resolution 12

13 Flexible frame design and radio resource management As indicated earlier, one of the key targets is to enable the frame to be able to carry allocations with different configurations optimized for the respective service/link/device characteristic. This includes aspects of: - Radio resource management - provision of a set of resource block configurations (from ultra-short for low latency services to very long for e.g. coverage extension) - Waveform design - Allow signal transmissions with less tight time/frequency alignments - Enable the possibility to adjust the subcarrier spacing per allocation - Optimize the filter characteristic according to the respective connection - Control channel design (DL) - Avoid shared control channels - Data and its related control should share the resources In-resource Control Channel (CCH) with downlink scheduling grant. Downlink data payload CCH content summary: - UE identifier - PHY configuration for data payload. - HARQ information - MIMO information 13

14 Extension of the FANTASTIC-5G air interface to specialized scenarios and extreme cases Phase 1 of horizon2020 is focused on designing a highly flexible air interface to support the anticipated heterogeneity forseen in 2020 Follow up actions apply the outcomes of phase 1 to specialized scenarios and extreme cases - Satellite communications - increased round-trip-time requires adapting feedback loops (e.g. HARQ) - different channel conditions possibly requires adaptations on waveform and frame design - low ARPU areas - ultra low-cost to be emphasized, e.g. by cutting down the system to only essential functionalities and by higher intersidedistances - Industry Extreme case of MCC - Emergency mode - Fall back mode using (non-network) controlled D2D and BMS features (e.g. to broadcast relevant info about rescue plans) Phase 1 design targets to ensure forward compatibility to those variants/adaptations. 14

15 Thanks! Questions? 15