5G Positioning for connected cars

Transcription

1 5G Positioning for connected cars (mmw) 5G introduction Mathematical model of 5G-mmW positioning Mutiple aspects of the achievable error Estimation principle June 2018 Summer school on 5G V2X communications Giuseppe Destino, CTR, King s College London

2 5G vs LTE-A (4G) 5G New carrier frequencies (sub-6 GHz and mmw) Beam-based communication New radio-access procedures New communication models Location-awareness

3 Com Range Frequency Radio-frequency map 0.8 GHz 6 GHz 100 GHz GHz 4G GHz 5G GHz 5G GHz 5G GHz 5G 10Km 200m 10m 1m Wide range Medium range Very-short urllc embb mmtc

4 mmw sparse channel model (Physical) Representation with 4 physical dimensions Sparse due to high carrier frequency

5 5G positioning: what is new Single BTS approach Use channel sparsity Use direction and distance infromation jointly Use a geomteric model to exploit one-bounce link

6 Positioning technology landscape Availability remote rural suburban city Satellite positioning A-GNSS (GPS, GLONASS, ) CI / E-CID based methods in-door BLE/UWB WiFi 1m km 3km 10km Accuracy

7 5G Positioning: mathematical model Estimate location and rotation of the MS Location-based channel parameterisation

8 When can we do positioning Positioning using beam sweeping Positioning using reference signals Inital access Sweeping multiple directions allows channel discovery No overhead (or reduced is specific pilots are needed) Periodically to allow location tracking Benefit Location-awareness prior communications

9 Signalling for positioning Beam Training DATA DL UL UE collects and processes signals over multiple beams gnb collects and processes signals over multiple beams Received OFDM signal at the m-th RF chain

10 Analysis of the Position FIM More information by increasing the subcarrier spacing More information by using edge-band subcarriers Bearing information depends on the sensitivity of the beampattern in angle domain AoA and AoD are coupled No derivations

11 Tool for performance analysis : CRLB FIM for channel parameters Apply variable transformation Compute the CRLB from the inverse of J η G. Destino, H. Wymeersch, On the Trade-off Between Positioning and Data Rate for mm-wave Communication, in IEEE International Conference on Communications Workshops, 2017

12 Position-rotation error bound DL mode LoS link provides 3 types of information AoA information: position-rotation dependent AoD information: position dependent Ranging: position dependent NLoS link provides a combined infromation AoA-Ranging information: position-scatter-rotation dependent The FIM of Position-Rotation in rank 3 in 2D, therefore position-rotation is feasible with 1 LOS + N >= 0 NLOS N >=3 NLOS R. Mendrzik, et All., Harnessing NLOS Components for Position and Orientation Estimation in 5G mmwave MIMO, arxiv 2017

13 Pilot signals Distributed Center-localised Edge-localised

14 Ranging error Scenario: AoA = 0 deg, AoD = 180 deg, d = 100m, 2 RF chains, 16 ULA

15 Ranging error 2 RF chains Impact of array gain Scenario: AoA = 0 deg, AoD = 180 deg, d = 100m, 2 RF chains, 16 ULA, orthogonal beams

16 Bearing error Due to the derivative of the beamforming Scenario: AoA = 0 deg, AoD = 180 deg, d = 100m, ULA, orthogonal beams

17 Achievable localisation error

18 Achievable localisation error

19 Requirements for 5G positioning AoA and AoD information Multiple beamforming to acquire information about AoA, AoD LOS / LOS and >1 NLOS / >= 3 NLOS Narrow beams for high SNR and high AoA/AoD resolution Spread pilots for high delay resolution

20 Sweeping strategy Exhaustive search Hierarchical search

21 Impact of beam training searching strategy Exhaustive search Information is acquired when main beam or sidebeam hit the LOS ray Resource consuming Beam-codebook dependent Hierarchical search Infromation is acquired at each step of the search High accuracy can be achieved as beams point to the right direction Time efficient

22 Trade-off: Rate vs Accuracy Hierarchical search: more time efficient but more sensitive to noise Exhaustive search: more robust to noise, trade-off between rate and positioning accuracy

23 Rate-PEB joint resource optimization User rate Time sharing optimisation

24 Position-estimation approach A. Shahmansoori, G. E. Garcia, G. Destino, G. Seco-Granados and H. Wymeersch, "Position and Orientation Estimation Through Millimeter-Wave MIMO in 5G Systems," in IEEE Transactions on Wireless Communications, vol. 17, no. 3, pp , March 2018

25 Subcarrier N Subcarrier 2 Subcarrier 1 Structured sparsity Angular domain Signals over multiple subcarriers share the same angular information Signals over multiple MIMO channels share the same time information

26 Two-step estimation in a nutshell Step1: Sparse estimation with common support Exploit common spatial-sparsity across carriers CS technique for common support model Estimate delay and channel gain per path Step2: Refinement of the channel parameters SAGE: per path refine the channel parameters using a successive cancellation method Step3: Non-linear mapping to location Solve non-linear least square problem

27 5G GNSS hybrid solution Sat1 Sat2 ψ 1 BTS θ 1 θ 2 ψ 2 -θ MS 0 -α On going work!

28 Reading 1. H. Wymeersch, G. Seco-Granados, G. Destino, D. Dardari and F. Tufvesson, "5G mmwave Positioning for Vehicular Networks," in IEEE Wireless Communications, vol. 24, no. 6, pp , Dec A. Shahmansoori, G. E. Garcia, G. Destino, G. Seco-Granados and H. Wymeersch, "Position and Orientation Estimation Through Millimeter-Wave MIMO in 5G Systems," in IEEE Transactions on Wireless Communications, vol. 17, no. 3, pp , March G. Destino and H. Wymeersch, "On the trade-off between positioning and data rate for mm-wave communication," 2017 IEEE International Conference on Communications Workshops (ICC Workshops), Paris, 2017, pp J. Saloranta, G. Destino and H. Wymeersch, "Comparison of different beamtraining strategies from a rate-positioning trade-off perspective," 2017 European Conference on Networks and Communications (EuCNC), Oulu, 2017, pp G. Destino, J. Saloranta, H. Wymeersch and G. S. Granados, Impact of Imperfect Beam Alignment on the Rate-Positioning Trade-Off, 2018 IEEE Wireless Communications and Networking Conference (WCNC): Special Session Workshops