(some) Device Localization, Mobility Management and 5G RAN Perspectives

Transcription

1 (some) Device Localization, Mobility Management and 5G RAN Perspectives Mikko Valkama Tampere University of Technology Finland December 16th, 2016

2 TAKE-5 and TUT, shortly Our work will primarily focus on WP3 and WP4 WP3 WP4 energy efficiency, mobility and localization related topics primarily with 5G RAN focus massive MTC, ultra-reliable MTC, multi-radio connectivity front-haul, back-haul, self-backhauling technologies all with 5G RAN focus TUT based PIs: Mikko Valkama, Sergey Andreev, Yevgeni Koucheryavy Contact: 2

3 WP3: 5G Radio Networks and Localization 5G networks, substantially enhanced positioning accuracy (3GPP TR & TR , 5GPP vision and white papers, NGMN 5G white paper, 5G Forum white papers, etc.) E.g., 1 meter accuracy outdoors substantially better compared to current networks (e.g. LTE OTDOA, UTDOA) Lots of new verticals calling for this, e.g. self driving cars, robots, drones, intelligent traffic systems, traffic management, Technical enablers dense networks, antenna arrays, wide bandwidth, multicarrier waveforms, short radio frames/sub-frames, frequent UL pilots

4 WP3: 5G Radio Networks and Localization Developed novel cascaded Extended Kalman Filter based solutions for High-efficiency ToA and DoA estimation and tracking in individual access nodes, based on UL reference signals Corresponding position estimation and tracking, as well as clock offset estimation and tracking in a central node Facilitates also mutual synchronization of the access nodes with mutual clock offsets

5 Example 1: localizing and tracking a moving car See demo at: 50 m Housing block Fig. Madrid grid Streets Access nodes: density ~ 50m* UN in LoS with two closest ANs Random route, v = km/h Assume ANs are mutually unsynchronized, UNs have clock offsets ANs have 2D concentric circular antenna arrays consisting of 9 cross-dipoles Uplink reference signal TX power = +3 dbm OFDM waveform, B = 100MHz, fc = 3.5GHz Detailed ray-tracing based propagation modeling 5

6 Example 1: localizing and tracking a moving car See video at: UL reference signals processed and EKFs updated only once per 100ms 6

7 Example 2: localizing and tracking a flying drone See demo at: A flying drone in LoS with two closest ANs Random flying route with max velocity of 50 km/h, including also landings and take-offs Antenna models, radio frame, etc. similar to the previous example Detailed ray-tracing based propagation modeling 7

8 Example 2: localizing and tracking a flying drone See video at: UL reference signals processed and EKFs updated once per 100ms 8

9 WP4: wireless self-backhauling We have been studying three different options for performing the self-backhauling while serving the s in the uplink (UL) and downlink (DL) Half-duplex scheme (classical, for refence) Full-duplex scheme Relay-type scheme Next, these different schemes are shortly described

10 Half-duplex backhauling Time slot 1 Time slot 2 Access node Backhaul node Access node Backhaul node The basic scheme, where transmission and reception are divided in time No interference between s, and essentially no interference at all assuming good beamforming

11 Inband full-duplex backhauling Access node Backhaul node In the full-duplex scheme, all transmission and receptions are done simultaneously High efficiency, but also complex interference mechanisms

12 Relay-type backhauling Time slot 1 Time slot 2 Access node Backhaul node Access node Backhaul node A relay-type scheme combines the good sides of half-duplex and full-duplex schemes less interference sources, while the full-duplex capability is still leveraged in the access node

13 Resource allocation: Optimizing the transmit powers Careful allocation of the different transmit powers is crucial to control the interference, and to improve the energy-efficiency This can be done in different ways, such as by mazimizing the sum-rate or ensuring a minimum Quality-of-Service (QoS) Here, the results are provided for a case where the transmit powers are minimized subject to a QoS requirement in the form of minimum per spectral efficiency Minimum data rate requirements for UL and DL 13

14 Example results Parameter Value Number of access node TX/RX antennas 200/100 Number of DL/UL s 10/10 Number of DL/UL backhaul streams 12/6 DL/UL rate requirement, per 8/2 bps/hz The optimized TX powers are calculated Cell radius Distance to the backhaul node Center-frequency 50 m 75 m for large amounts of randomly dropped s in the network 3.5 GHz The cumulative distribution functions of the transmit powers can be compared between the different schemes This shows which of the schemes is capable of fulfilling the same QoS requirements with the best energy-efficiency (lowest transmit powers)

15 Example result 1 The full-duplex (FD) scheme achieves significantly lower transmit powers both in the AN and in the s than the relaytype (RL) or half-duplex (HD) schemes However, for some of the positions, it cannot fulfill the QoS requirements with any finite transmit power (since the CDF saturates to a value less than 1)

16 Example result 2 When investigating the total TX power usage (UL+DL), the FD scheme also outperforms the other solutions However, with less SI cancellation, the probability of not fulfilling the QoS requirements with any finite transmit power is larger Also, the relay scheme starts to perform rather poorly when the amount of SI cancellation is 110 db or less

17 Inhouse prototype device for inband full-duplex Fully operational full-duplex demonstrator developed and up and running at TUT Operates at 2.4 GHz ISM band, supports up to 200 MHz instantaneous BW Contains advanced RF selfinterference cancellation as well as novel digital selfinterference cancellation solutions, all in real time Original transmit data (x(n)) PA TX Pre-computed NI PXIe-7972R with Kintex-7 FPGA NI 5791 RF transceiver LPF L x(n) x(n) 2 x(n) x(n) P-1 x(n) RF canceller RX Orthogonalization h 1 h 2 h P Σ Σ Σ LMS filter weight update Cancelled signal

18 Example RF measurement results Live measurements incorporating also Aalto-based back-to-back relay antenna*, 80 MHz instantaneous BW at 2.4 GHz More than 100 dbs of measured TX-RX isolation * D. Korpi, M. Heino, C. Icheln, K. Haneda, and M. Valkama, "Compact inband full-duplex relays with beyond 100 db self-interference suppression: Enabling techniques and field measurements," IEEE Transactions on 18 Antennas and Propagation, accepted, to appear, 2016.

19 WP4: Massive and Ultra-reliable MTC 19 Objectives New MAC, RRM and protocol solutions to enable novel 5G use cases related to both massive and critical MTC. Support of ultra-reliable MTC applications and services. Challenges Understanding channel behaviour (propagation) to support critical and massive connectivity communications in 5G-grade MTC scenarios. Propose models and solutions able to enable ultra-reliable low latency communications.

20 75 m 20 5G-grade IoT research on factory automation Since statistical channel models are not suitable for factory environments, a comparison between real and statistical path loss measurements has been investigated by using our in-built Ray-based (RL) and system-level (SLS) tools. The conclusion is the impossibility to generalize a path loss formula for any indoor scenario. Thus, a proper one has to be achieved in accordance to the environment considered. 150 m Transmitter 1 Transmitter 2 Machinery area Office area

21 SNR DL Heatmaps 21 Ray-based simulator System-level simulator (SLS)

22 Pathloss assessment 22 Statistical channel models are only suitable for scenarios in which they were made. The gap between deterministic and real models is very important. TUT Ray-based tool is useful to characterize practical channel propagation. A deterministic path loss formula can be obtained through extensive simulations. In order to conduct a comprehensive system-level analysis, propagation behaviour needs to be understood correctly. In doing this, our RL tool provides powerful instruments to have a complete understanding on wireless channel propagation in factories of the future (i.e., Industry 4.0)

23 WP4: Multi-RAT Integration Reliability in Multi-RAT networks research track Objectives: Study options for Multi-RAT integration on different layers. Improve reliability in Multi-RAT networks by using simultaneous connections to various RATs. Challenges: Finding application independent solutions Balancing energy efficiency with reliability/performance gains

24 TUT testbed setup TUT testbed with Ericsson pico base stations, which are connected to the Aalto EPC: has both LTE and WiFi interfaces on at the same time, and can send the same data duplicated over both links. Duplication currently works only on application layer

25 Example scenario concept Application used is simple echo server. User is moving out of WiFi coverage, so starts sending data over LTE as well

26 Future work Consider additional scenarios: Mission critical: Constantly send over all interfaces for maximum reliability. IoT: Balance energy efficiency with reliability gains. Test with real applications Prototype demo