Ultra-Reliable Wireless Communications: Building Blocks and Models. Petar Popovski Aalborg University Denmark

Transcription

1 Ultra-Reliable Wireless Communications: Building Blocks and Models Petar Popovski Aalborg University Denmark

2 outline the state of wireless, MTC (M2M) and 5G URLLC short packets wireless channel modeling perspective interface diversity

3 the state of wireless, M2M, and 5G

4 historical impact of mobile wireless from World Development Report 2016: Digital Dividends

5 wait, mobile phone yes, but not electricity? solar kiosk franchisee on the streets of Kigali, Rwanda from

6 where does wireless technology move the business of transporting data bytes from A to B has become less lucrative instead it became important what one does with the data that has been transported

7 wireless: why are we not there yet driverless cars and assisted driving remote monitoring pictures from Clipart Panda and Shutterstock

8 how is MTC positioned in 5G M2M communication = MTC (Machine-Type Communication) Extended MBB IMT 2000 & IMT-Advanced MBB Massive MTC Ultrareliable MTC IMT for 2020 and beyond Additional new use cases Massive MTC Low cost, low energy TEXT Small data volumes Massive numbers Ultra reliable Very low latency Very high availability Ultra-reliable low latency MTC Smart city, infrastructures and objects (massive devices, e.g. sensors and actuators) Autonomous vehicle control Smart grid Factory cell automation from 5G wireless and mobile communication technology"

9 how is MTC positioned in 5G MTC stands for 2/3 of 5G the official ITU-R terminology on the 5G generic services: Enhanced mobile broadband (embb) Massive machine-type communications (mmtc) Ultra-reliable and low-latency communications (URLLC) challenges in designing the technology for each service coexistence of services and network slicing

10 data rate new modes: massive and ultra-reliable access data rate what today s wireless systems cannot do 1 Mbps from 100 devices or 10 kbps from devices 100 Mbps 95% of the time or 100 kbps % of the time access protocol limit reliability limit for control information # devices error probability

11 ultra-reliable low latency communication URLLC

12 ultra-reliable communication (URC) as an ecosystem URC is about turning wireless into a reliable commodity and it is commodity when the engineers using wireless ask we have ultra-reliable wireless, can we think of an application that can use it? figure from

13 CDF CDF design target for URLLC Normal reliability (e.g. mobile Internet) Metrics: peak, median, 95-percentile Exploit opportunities of favorable conditions Ultra-reliable low latency communication Metric: percentile Focus on worst-case 100% 100% Guaranteed delay limit latency latency

14 building blocks for URLLC transmission of short packets high diversity restricted opportunities for retransmission lean protocol design with respect to latency requirements in ms new precise channel models and statistically correct description of rare events capture the events with probability 10E-8

15 ultra reliable transmission of short packets Philippe Petit,

16 anatomy of a digital connection Alice sends data to Bob but Alice also must send metadata or control data to Bob let Alice send header of H bits using m channel uses data of D bits using n channel uses

17 goodput, metadata included [bps] high-speed wireless ultra-reliable

18 common assumption and design approach the metadata size negligible compared to the data use heuristic methods to send metadata use sophisticated wireless to send data

19 the relationship of data and metadata this relationship is changed in short packets suddenly it matters a lot how we send preambles, CRC, node address, etc. transition similar to the one from classical to quantum physics

20 Shannon s communication model very general model of information systems with random disturbances

21 communication-theoretic model for short packets

22 fundamental theory of finite blocklength transmission at short blocklengths there is a penalty that keeps the rate away from capacity. AWGN SNR 0 db

23 illustration of gain in reliability low SNR 10 bytes control information 10 bytes data same amount of channel uses probability of error 10e-3 probability of error 10e-6

24 frequency common assumption and design approach frequency M D M D time time the notion of frame in cellular systems should be revisited

25 connection between packetization and energy efficiency separated data and metadata useful for energy efficiency data for Bob, Carol turns off her receiver after the metadata Bob Alice Carol joint data and metadata Alice better coding of the metadata however, everybody decodes everything Bob Carol

26 observations basic tradeoff between energy efficiency and ultra-reliability departure from the common causal relationship metadata -> data low latency usually means sending with few channel uses (DoF) DoF can be increased in e.g. frequency or space

27 revisiting data and metadata: example of framed downlink transmission a BS has unicast packets to K devices standard approach is TDMA finite blocklength theory suggests to concatenate the packets and broadcast a single long data packet take-home: improve own reliability by decoding the data of the others

28 scenario downlink communication to K users a user is active and there is a packet for her with probability q. the message for each active user is drawn randomly from a set of predefined message sizes. metadata should inform about who is active the message size.

29 downlink framing conventional frame-based transmission with individual pointers

30 ptr 4 ptr 3 ptr 2 downlink framing UG 1 UG 2 UG 3 UG 4 m essage sizes m essages with messages wit h framing with user grouping and joint packet encoding

31 latency-energy tradeoff new tradeoff arises for short packets latency is minimized when all packets are jointly encoded; power is minimized when each packet is encoded separately. K=32 users

32 wireless channel modeling for ultra-reliable communication

33 how is this related to communication engineering a communication engineer models known unknowns y = h x+ z+i channel state noise interference URC requires a bounding the impact of the unknown unknowns opening for methods of risk analysis in wireless system design

34 wireless channel model behavior in ultra-reliable regime currently there is lack of experimental evidence for URC-relevant statistics of wireless channels initial analysis of common wireless channel models in URC regime block fading P R is the minimal SNR to decode data rate R the analysis reveals the URC-behavior: Pr P R P < L ε α P R P β

35 outdoor physical setup istic two-path model H TX H RX V TW = Gλ 4 d L e j 2 λ d L + Γ Gλ 4 d R e j 2 λ d R R is the direct/reflected path distance. two-wave model with equal amplitudes r = V TW = exp j φ represents one of the worst cases (Aalborg Uni) ultra-reliable wireless ITA, San Diego, Feb / 17

36 indoor physical setup indoor case dominated by diffuse components

37 2-wave, 3-wave and Rayleigh comparison

38 ultra-reliability using multiple communication interfaces

39 multi-interface transmission interface diversity 1 reliability: P(X x) Transmission errors Infrastructure failures P e transmission strategies cloning latency: x timeout latency distribution 2-out-of-3

40 interface diversity experimental results results based on lab measurements 1 day, 100 ms interval Wi-Fi achieves 10 ms for 90% of packets but 99% requires almost 100 ms cellular: LTE and HSPA also requires ~100 ms for 99% cloning (1 copy per IF) 99% at 25 ms % at 60 ms latency LTE HSPA Wi-Fi Cloning all 2-out-of-3 RTT l [ms] interface diversity can reduce latency and improve reliability by several orders of magnitude

41 final remarks

42 protocol-level view makes the challenge much bigger UE enb 1. Random Access preamble 2. Random access response 3. RRC Conn. Request 4.a. Contention resolution Access Attempt 4.b. RRC Conn. Setup 5. RRC Conn. Setup Complete 6. RRC Security Mode Command 7. RRC Security Mode Complete 8. RRC Conn. Reconfiguration Connection Establishment 9. RRC Conn. Reconf. Compl. 10. Small Data Payload LTE access for a short packet Source: Ciscus Sarasota

43 conclusions and outlook ultra-reliable wireless has the potential to profoundly change systems and devices essential: short packet transmission and the associated protocols proper wireless channel modeling use as many diversity possibilities

44 final words: the power of information sardine prices in three coastal markets in India (World Development Report 2016) now think about how many processes and business could be affected through Internet of Things

45 thanks Kasper F. Trillingsgaard Jimmy J. Nielsen Patrick C. F. Eggers