Principles of V2X Radio Interface Design Tommy Svensson Professor, - - PowerPoint PPT Presentation

• Dept. of Electrical Engineering, Communication Systems

Slide 1

Principles of V2X Radio Interface Design

Tommy Svensson

Professor, PhD, Leader Wireless Systems Department of Electrical Engineering, Communication Systems Group Chalmers University of Technology tommy.svensson@chalmers.se

• Dept. of Electrical Engineering, Communication Systems

Slide 2

Outline

• Why Cellular-Assisted V2X?
• Designing the 5G V2X Radio Interface
• Integrated Moving Networks
• Conclusions

• Dept. of Electrical Engineering, Communication Systems

Slide 3

V2X Basics

• Vehicle-to-everything (V2X) communication: Passing of information from a

vehicle to any entity that may affect the vehicle, and vice versa.

• A vehicular communication system incorporating specific types of

communication – V2I (Vehicle-to-Infrastructure) – V2V (Vehicle-to-vehicle) – V2P (Vehicle-to-Pedestrian) – V2D (Vehicle-to-device) – V2G (Vehicle-to-grid).

• The main motivations for V2X are safety and energy savings.
• V2X communication was originally based on WLAN technology forming a

vehicular ad-hoc network as two V2X senders come within each other’s range.

• “Hence it does not require any infrastructure for vehicles to communicate,

which is key to assure safety in remote or little developed areas.”

• “WLAN is particularly well-suited for V2X communication, due to its low
• latency. It transmits messages known as Common Awareness Messages (CAM)

and Decentralised Notification Messages (DENM) or Basic Safety Message (BSM). The data volume of these messages is very low. The radio technology is part of the WLAN IEEE 802.11 family of standards and known in the US as Wireless Access in Vehicular Environments (WAVE) and in Europe as ITS-G5.”

Source: Wikipedia

• Dept. of Electrical Engineering, Communication Systems

Slide 4

Functions in Intelligent Transportation Systems (ITS)

• Forward collision warning
• Lane change warning/blind spot warning
• Emergency Electric Brake Light Warning
• Intersection Movement Assist
• Emergency Vehicle Approaching
• Road Works Warning
• Platooning
• …

Source: Wikipedia

• Dept. of Electrical Engineering, Communication Systems

Slide 5

Intelligent Transportation Systems (ITS) – Why Long Information Horizon Matters

https://www.youtube.com/watch?v=iHzzSao6ypE&feature=youtu.be

• Dept. of Electrical Engineering, Communication Systems

Slide 6

Wireless Communications Everywhere and with “Everything”

• Dept. of Electrical Engineering, Communication Systems

Slide 7

Increasing Need for Mobile Broadband

• Dept. of Electrical Engineering, Communication Systems

Slide 8

Original Motivation cont.

• A larger number of mobile users will be vehicular

Home access Internet Office access Internet On-road access Internet USA 37.8% 19.6% 42.6% UK 45.6% 17.8% 36.6% Germany 43.4% 15.3% 41.3% France 33.1% 21.7% 45.2% Italy 39.6% 21.4% 39.0% South Africa 48.6% 21.4% 30.0% Mexico 28.2% 27.6% 44.2% Brazil 36.7% 24.7% 38.6% Korea 33.7% 31.7% 34.6% India 45.9% 30.4% 23.7% China 30.1% 32.7% 37.2%

Source: Cisco VNI Mobile, 2011

METIS | 5G V2X Communications - Summer School | 2018-06-11 | Page 9

TC6: Traffic jam

Provision of public cloud services inside vehicles during traffic jams due to the sudden increase in the capacity demand – Traffic volume: 480 Gbps/km2 – User data rate: 100/20 Mbps in DL /UL with 95% availability

Test Cases related to Moving Networks

TC10: Emergency communications

Basic communications in a place where little mobile or wireless network infrastructure exists, e.g. due to a natural disaster. – Battery lifetime: 1 week (with today’s battery technology) – Availability: 99.9% victim discovery rate – Destroyed or unreliable NW infrastructure

Survivor with working UE Dead victim but with working UE Temporary nodefor rescue

• perations

macro cell remaining aliveafter earthquake Destroyed building

METIS | 5G V2X Communications - Summer School | 2018-06-11 | Page 10

TC7: Blind spots

The ubiquitous capacity demands in blind spots, such as rural areas with sparse NW infra- structure or in deeply shadowed urban areas. – User data rate: 100/20 Mbps in DL/UL – Energy efficiency: 50% / 30% reduction for UE / infrastructure

TC8: Real-time remote computing for mobile terminals

Remote computing services, e.g., augmented reality service, on-the-go at higher speeds. – User data rate: 100/20 Mbps in DL /UL – Latency: Less than 10 [ms] with 95% reliability – Mobility: Up to 350 km/h

Test Cases related to Moving Networks

METIS | 5G V2X Communications - Summer School | 2018-06-11 | Page 11

TC11: Massive deployment of sensors and actuators

Small sensors and actuators that are mounted to stationary or movable objects and enable a wide range of applications – Energy efficiency: 0.015 μJ/bit for 1 kbps data rate – Protocol efficiency: 80% at 300,000 devices per access node – Availability: 99.9%

TC12: Traffic efficiency and safety

Cooperative intelligent traffic systems (C-ITS) for road safety and traffic efficiency – Latency: Less than 5 [ms] for 99.999% – Detection range: up to 1 km – Availability: ~100%

Test Cases related to Moving Networks

METIS | 5G V2X Communications - Summer School | 2018-06-11 | Page 12

“Moving Networks” refers to novel concepts that focus on moving and/or nomadic network nodes & terminals.

• Cluster #1:

Mobility-robust high-data rate comm. links

• Requirement: High-data Rate, Low Latency
• Relaying inside vehicles is not the only focus

Moving Networks in the METIS project

• Cluster #2: Flexible network deployment based
• n nomadic network nodes
• Requirement: High Data Rate
• Relaying inside vehicles is not considered here!
• Cluster #3: V2X communications
• Requirement: Low-Medium Data-Rate,

Low Latency, High Reliability

• Dept. of Electrical Engineering, Communication Systems

Slide 13

General Motors' EN-V concept

https://www.youtube.com/watch?v=0tiHwzGsotA

• Dept. of Electrical Engineering, Communication Systems

Slide 14

Phantom Auto's Remote Driving

https://www.youtube.com/watch?v=HlGqYFclKqU

On the 5GCAR Use Cases

• Cooperative safety: achieved by exchanging the

information about detection of the presence of road users

• Network assisted vulnerable pedestrian

protection

• Cooperative perception: perception extension

is built on the basis of exchanging data from different sources, e.g., radars, laser sensors, stereo-vision sensors from on-board cameras

• See-through
• Cooperative maneuver: sharing local awareness

and driving intentions and negotiating the planned trajectories

• Lane merge

15

On the 5GCAR Use Cases

• Remote driving: control the different actuators of

the car (steering wheel, brake and throttle) from

• utside the vehicle through wireless

communication

• Remote driving for automated parking

Further information: https://5gcar.eu/.

• Autonomous navigation: construction and

distribution of real-time intelligent HD map

• High definition local map acquisition

16

• Dept. of Electrical Engineering, Communication Systems

Slide 17

Designing the 5G V2X Radio Interface

– Towards a Reliable High Capacity Infrastructure Interface

• Robust
• High capacity
• Low latency
• Support multicast/broadcast
• Efficient also for small packets

Ultra-Reliable Low-Latency Communication (URLLC).

Slide 18 Event: 5G V2X Communications - Summer School Date: June 11, 2018

• Utilize the information on channel

variability in order to achieve statistical multiplexing gain on the time-frequency (and space) resources

• Use link adaptation and more or less
• pportunistic scheduling (under QoS and

certain fairness criteria)

• Spectral efficiency increases with number
• f users (multi-user diversity gain).
• Collect channel quality information to scheduler

by using channel prediction (SINR prediction) and send this information over a feedback channel (FDD)

• Control

interference by keeping users

• rthogonal within

cell (no spreading), and by using regulated frequency bands

Adaptive Transmission

Adapt to the Frequency-Selective Small-Scale Fading of the Channel

Slide 19 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Basic Radio Resource Units: Chunks and Chunk Layers

Tchunk BWchunk nsymb OFDM symbols nsub sub- carriers Chunk Chunk time frequency

Layer 1 Layer 2 Layer 3 Layer 4

layer a) b)

a) Physical channel structure and chunks b) Chunk layers obtained by spatial re-use.

The channel is essentially flat within a chunk.

Duplex guard time 8.4s 390.62 KHz 0.3456 ms for 1:1 asymmetry 0.3456 ms chunk duration 15 OFDM symbols 12 OFDM symbols Time Time f f 8 subcarriers 8 subcarriers FDD mode TDD mode

96 symbols

312.5 KHz

120 symbols

Slide 20 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Short Frame Structure Enabling Frequency-Adaptive Transmission

DL UL DL

O O O O O O O O C C C C C C C C D D P U U P D D P U U P D D P U U P D D P U U P

• 1. DL control symbols:

Report which present chunks belong to which flows

• 2. DL pilot symbols:

Used for channel prediction Used for channel estimation

• 3. UL control symbols:

Report which next UL chunks appointed to which uplink flows

• 4. Pilot symbols:

Used for coherent detection And updating predictor states

UL

O O O O O O O O C C C C C C C C

• 6. UL overlapping pilot symbols:

Used for prediction

• 5. DL control feedback symbols:

Carry DL channel prediction report Dl prediction horizon: 2.5 x 0.3372ms=0.843ms UL prediction horizon: 2.5 x 0.3372ms=0.843ms

• Note, example! For final WINNER II Reference Design: See deliverable D6.13.14

Short prediction horizon enables vehicular speeds!

Slide 21 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Prediction Horizon and SINR Limit at 5 GHz FDD Downlink



Jake’s Doppler spectrum assumed.



12.5 dB, 0.273 6 dB, 0.195 <0 dB, 0.117 70 km/h 50 km/h 30 km/h



Prediction error target: NMSE=0.15.



vD/

Prediction horizon:

Slide 22 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Adaptive Coding and Modulation (MI-ACM)

chunk Nch

M U X 1

P

channel encoder

1 2 4 6 8

M U X 2

1 nf

p1

1 nf

pNch

IFFT +CP Adaptive bit loading

K·Ncw R N·Ncw Ncb Npad

Coding with adaptive puncturing AWGN FFT

• CP

1 nf 1 nf

D E M U X 2

1 2 4 6 8

D E M U X 1

APP demapping

P-1

channel decoder Npad

Decoding h(n)

chunk 1

Tx Rx

chunk Nch chunk 1

• Local chunk decision on adaptive M-QAM modulation
• Common FEC code over the chunks

– Code rate calculated using Mutual Information based averaging

• ver the M-QAM modulated chunks

– Common outer puncturing

• Low computational complexity
• Not restricted to a specific coding scheme

Slide 23 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Adaptive Transmission

• Frequency-channel-dependent scheduling & fast link adaptation

– Spectrally efficient transmission – Adapt per chunk layer to small-scale fading

• Non-frequency adaptive transmission (for fast terminals, multicast, ...)

– Robust transmission – Adapt per frame to shadow fading and path loss

Link adaptation Fast scheduling Fast retransmissions with soft combining

Slide 24 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Localized and Dispersed Resource Allocation

B-IFDMA: Block Interleaved Frequency Division Multiple Access B-EFDMA: Block Equidistant Frequency Division Multiple Access

• Both frequency-adaptive

transmission and non- frequency-adaptive transmission benefit from frequency diversity

• Low delay requirements

and availability of large system bandwidth makes

frequency multiplexing

• f resources most beneficial.
• Facilitated by large system

bandwidth per operator

Frequency

Non-frequency adaptive (B-IFDMA or B-EFDMA)

Time

chunk chunk

User 2: User 3: User 4:

chunk

Frequency adaptive (TDMA/OFDMA) User 5: User 6: User 1:

Slide 25 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Resource . Scheduler . Packet . Processing IFFT, beamforming, pulse shaping Tchunk Time Frequency PHY CQI, CSI Buffer levels CQI errors Antenna 1 .... Antenna N Non -frequency -adaptive Resource Scheduler Resource Scheduler Map on dispersed chunks Map on optimal chunks

• Scheduled

Mapping

• Link

adaptation RLC SDUs MAC RLC Frequency-Adaptive Segmentation/Concatenation FEC Coding HARQ Retransmission Buffers

MAC Support of Frequency-adaptive and Non-frequency-adaptive Transmission

Two distinct transmission modes:

• frequency-adaptive

transmission: adapt to frequency-selective small- scale fading

• non-frequency-adaptive

transmission: rely on diversity Fast Resource scheduling and Link adaptation requires

• good channel knowledge,

which requires

• predictable channels and

interference environment Fast adaptation of transmission mode is needed and is enabled by:

• common channel coding

and Hybrid-ARQ approach

• fully controlled by MAC

Slide 26 Event: 5G V2X Communications - Summer School Date: June 11, 2018

Conclusions on WINNER Adaptive Transmission

• Two basic transmission modes are supported in the WINNER System

Concept:

– Frequency-adaptive transmission – Non-frequency-adaptive transmission

• Significant gains from multi-user scheduling and frequency-adaptive

link adaptation have been obtained

• Vehicular user speeds can be supported in wide area scenarios by

using channel prediction

• Prediction accuracy depends especially on

– SINR – Doppler spectrum and – Prediction horizon in wavelengths (which depends on prediction horizon in time, user speed and carrier frequency)

• Error rate targets can be fulfilled also in the presence of SINR

prediction errors (non-perfect CQI)

• Outer channel coding with per chunk adaptive modulation combines

multi-user scheduling gains with near optimal link level performance without need for power loading

• M. Sternad, T. Svensson, G. Klang, ”WINNER MAC for Cellular Transmission,” Proc. 15th IST

Mobile & Wireless Communications Summit, Myconos, Greece, 2006.

• Dept. of Electrical Engineering, Communication Systems

Slide 27

How to Improve CSIT at High Speed?

IMT Advanced requirements at high speed: 10 times lower requirements at high speed!

• Dept. of Electrical Engineering, Communication Systems

Slide 28

The Predictor Antenna Concept

v

• Antenna elements of an antenna array may act as predictors for next antenna

element.

• Can be combined with conventional prediction based on past data to boost

performance further. Challenges: Possible decorrelation due to

› Effects of the moving vehicle on the standing wave pattern › Non-equal scattering environment around the antennas › Mutual electromagnetic coupling of antennas (Vaughan 1991)

• Dept. of Electrical Engineering, Communication Systems

Slide 29

Averages and 5% and 95% percentiles for all subcarriers at three test runs

0.5 1 1.5 2 2.5 3 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 Antenna distances [] Correlation LOS average NLOS average LOS 5%-percentile NLOS 5%-percentile LOS 95%-percentile NLOS 95%-percentile

• Two antennas in urban environment (line-of-sight, non-line-of sight).
• 20 MHz OFDM downlink at 2.68 GHz, measurements at vehicular velocity 45-50 km/h
• M. Sternad, M. Grieger, R. Apelfrojd, T. Svensson, D.

Aronsson, A. Belen Martinez “Using “Predictor Antennas” for Long-Range Prediction of Fast Fading for Moving Relays,” IEEE Wireless Communications and Networking Conference (WCNC), 2012.

Refined measurements show even better results, especially for NLOS conditions!

Boosting the MN Backhauling - Predictor Antenna

• Dept. of Electrical Engineering, Communication Systems

Slide 30

Refined Research Results

0.5 1 1.5 2 2.5 3 0.7 0.75 0.8 0.85 0.9 0.95 1 Antenna distances [] Correlation Monopoles, with metal sheet LOS average NLOS average LOS 5%-percentile NLOS 5%-percentile LOS 95%-percentile NLOS 95%-percentile 0.5 1 1.5 2 2.5 3 0.7 0.75 0.8 0.85 0.9 0.95 1 Antenna distances [] Correlation Dipoles, with metal sheet LOS average NLOS average LOS 5%-percentile NLOS 5%-percentile LOS 95%-percentile NLOS 95%-percentile LOS w/ ghost antennas NLOS w/ ghost antennas

With monopol antennas: With dipole antennas:

• Dept. of Electrical Engineering, Communication Systems

Slide 31

Predictor Antenna Experimental Results

• Two antennas in urban environment (line-of-sight, non-line-of sight).
• 20 MHz OFDM downlink at 2.68 GHz, measurements at 45-50 km/h

With Antenna embedded pattern compensation:

• N. Jamaly, R. Apelfrojd, A. Belen Martinez, M. Grieger, T. Svensson, M. Sternad and G. Fettweis, ”Analysis and Measurement of Multiple

Antenna Systems for Fading Channel Prediction in Moving Relays,” EuCAP’2014, April 2014, Haag, The Netherlands.

• M. Sternad, M. Grieger, R. Apelfrojd, T. Svensson, D. Aronsson, A. Belen Martinez “Using “Predictor Antennas” for Long-Range Prediction
• f Fast Fading for Moving Relays,” IEEE WCNC, Paris, 2012.

› Prediction horizon beyond 3λ seems feasible => x10 better – enables closed loop schemes at high speed >1 GHz !

v

• Dept. of Electrical Engineering, Communication Systems

Slide 32

Application to Massive MIMO Downlink

predictor antenna a) RS b) SRTA uplink downlink Beamforming mispointing! Beamforming reaches the target! vt

( ) ( )

a

vt v vt v  Reference System Use of predictor antenna

• D. Thuy, M. Sternad and T. Svensson, “Adaptive Large MISO Downlink with Predictor Antenna Array for Very

Fast Moving Vehicles,” ICCVE’2013, Dec 2013, Las Vegas, USA.

• Dept. of Electrical Engineering, Communication Systems

Slide 33

Robustness

• Link performance - Robustness metric: Block Error Rate (BLER)

SRTA-PI: PI is triggered for speeds >10kmph.

SRTA: Separate Receive and Transmit Antennas

• Dept. of Electrical Engineering, Communication Systems

Slide 34

Energy savings

Definition: eS= ESISO/EMISO.

• ESISO and EMISO: required

energy with SISO and MISO to achieve target BLER of 0.01 in 64QAM, with coding rate ¾.

• Link performance - Energy saving metric eS

• Dept. of Electrical Engineering, Communication Systems

Slide 35

Moving Relays/BSs/Cells

• Natural hot spots!
• Avoid vehicular penetration loss (VPL): 25-30

dB and more for >6Ghz

• Improve UE battery life time
• Enable closed-loop schemes for fast moving

users

• Onboard multi-service resource control

Challenges

• Backhaul design for moving relays/BSs/cells
• Resource allocation/partitioning for macro users and

moving cell users

• Interference coordination
• Handover with assured QoS

– Group user approach?

• Cooperation involving moving relays/base stations
• How to best serve outdoor users with moving base

stations

• Y. Sui, A. Papadogiannis, J. Vihriälä, M. Sternad, W. Yang, T. Svensson, IEEE “Moving Relay Nodes: A promising

solution to boost performace of vehicular users”, Special Issue IEEE Communications Magazine.

• Dept. of Electrical Engineering, Communication Systems

Slide 36

Moving Relay System

Outage Probability at 25dB VPL

200 400 600 800 1000 10

• 10

10

• 8

10

• 6

10

• 4

10

• 2

10 Outage Probability R < 2 bit/s/Hz at UE UE distance from BS [m]

Direct transmission simulated Direct transmission analytical MRN case simulated MRN case analytical FRN optimized simulated FRN optimized analytical FRN lower bound simulated FRN lower bound analytical

Ergodic Capacity at 25dB VPL

200 400 600 800 1000 5 10 15 20 25 30 35 Ergodic Capacity [bit/s/Hz] UE distance from BS [m]

Direct transmission MRN case FRN optimized FRN lower bound

• Dept. of Electrical Engineering, Communication Systems

Slide 37

Two Cell System Model

• Target: Vehicular UE
• Considered Metric: Outage Probability
• Schemes: Direct transmission (base line), Fixed Relay Node (FRN) and

Moving Relay Node (MRN) assisted transmission.

• A two cell deployment: Primary cell and Interfering Cell
• Sui, Y. ; Papadogiannis, A. ; Yang, W. et al. (2012). Performance Comparison of Fixed and Moving Relays under Co-

channel Interference, IEEE 4th Int. Workshop on Heterogeneous and Small Cell Networks (HetSNets), Globecom 2012.

• Dept. of Electrical Engineering, Communication Systems

Slide 38

Simulation Results - Outage Probability

• When VPL is 10 dB, direct

transmission and MRN assisted transmission gives almost the same OP at the UE

• Difference between MRN worst

case and MRN average case is barely noticeable due to the interference is attenuated twice by both vehicles.

• FRN is good at serving the UEs

around it.

Outage Probability at 10dB VPL

• Dept. of Electrical Engineering, Communication Systems

Slide 39

Simulation Results - Outage Probability

Outage Probability at 30dB VPL

• When VPL is 30 dB, MRN

assisted transmission significantly outperforms direct transmission.

• The contribution of FRN to

vehicular UE is barely noticeable due to its low transmit power.

• Dept. of Electrical Engineering, Communication Systems

Slide 40

Handover Optimization

• Short Time-to-trigger (TTT) lowers the power Outage

Probability.

• Longer TTT reduce the Ping-Pong handover rate
• Y. Sui, Z. Ren, W. Sun, T. Svensson and P. Fertl, “Performance Study of Fixed and Moving Relays for Vehicular

Users with Multi-cell Handover under Co-channel Interference,” ICCVE’2013, Dec 2013, Las Vegas, USA.

• Dept. of Electrical Engineering, Communication Systems

Slide 41

Moving Base Station (MBS) to form Moving Cells/Networks

• Full-duplex out-band

moving base stations (MBSs) forming a moving network (MN) inside vehicles

• Independent resource

allocation within vehicle

• Context awareness

buffering and scheduling

• Potential for caching

system

• Dept. of Electrical Engineering, Communication Systems

Slide 42

• METIS Test Case 2
• Calibration setup in D6.1
• Macro sectors at 800 MHz

with 20 MHz bandwidth

• Micro sectors at 2.6 GHz with

80 MHz bandwidth

• Moving Networks are out-

band and full-duplex

• Public transportation vehicles

enter the streets according to a Poisson distribution with a speed of 50 km/h.

System Model

• Dept. of Electrical Engineering, Communication Systems

Slide 43

Illustration of the Interference Situation

Moving network serving users in a vehicle Micro BS at street level, 2 sectors

• Dept. of Electrical Engineering, Communication Systems

Slide 44

SNR and SINR at Vehicular UEs (VUEs)

Cumulative distribution function of SNR and SINR at vehicular UEs (VUEs) and MN backhaul links at 15dB VPL

• Dept. of Electrical Engineering, Communication Systems

Slide 45

• Backhaul links of MNs
• Multi-antennas

MRC (Maximum Ratio Combining) when only the channel state information (CSI) of the desired signal can be obtained. IRC (Interference Rejection Combining) if CSI of both the desired and the interfering signal can be obtained.

• Access links of MNs
• Micro cells use Almost Blank Subframes (ABSs).
• Minimum ABS ratio is dynamically defined such that all

the access links can accommodate the traffic from their backhaul links.

ICIC Schemes Considered in this Study

• Dept. of Electrical Engineering, Communication Systems

Slide 46

VPL 15 dB

VUE Throughput

VPL 30 dB › Substantial throughput improvement for VUEs with MRN (MBS) using IRC

• Dept. of Electrical Engineering, Communication Systems

Slide 47

VPL 15 dB

Macro UE Throughput

VPL 30 dB › Minor throughput degradation for macro-UEs with MRN (MBS)

• Dept. of Electrical Engineering, Communication Systems

Slide 48

VPL 15 dB

Micro UE Throughput

VPL 30 dB › Throughput degradation for micro-UEs with MRN (MBS) due to ABS pattern, minor at high VPL.

• Dept. of Electrical Engineering, Communication Systems

Slide 49

• Moving relays/base stations are very useful in low

interference scenarios

• Interference management is very important for the use of

MNs in ultra-dense urban scenarios

• With advanced ICIC schemes, the use of MNs can improve

the experience of VUEs without significantly degrading the performance of outdoor UEs. High VPL is beneficial.

• Advanced backhaul links are the key to further boost the

performance of MNs.

Summary of Observations

• Y. Sui, I. Guvenc, T. Svensson, ”Interference management for moving networks in ultra-dense urban scenarios”, EURASIP

Journal on Wireless Communications and Networking 2015

• Y. Sui, A. Papadogiannis, J. Vihriälä, M. Sternad, W. Yang, T. Svensson, IEEE “Moving Relay Nodes: A promising solution to

boost performance of vehicular users”, Special Issue IEEE Communications Magazine.

• METIS D3.2 “First performance results for multi-node/multi-antenna transmission technologies”, April 2014
• METIS D4.3 “Final report on network-level solutions”, to appear Feb 2015
• Y. Sui, I. Guvenc, and T. Svensson, “On the Deployment of Moving Networks in Ultra-dense Urban Scenarios”, IEEE

International Conference on 5G for Ubiquitous Connectivity (5GU), Nov, 2014

• Dept. of Electrical Engineering, Communication Systems

Slide 50

Fully Integrated Moving Cells in Dense Small Cells Heterogeneous Network (HetNets)

With CSIT based on evolved Predictor antenna systems we can fully integrate moving base stations in a generalized HetNets

• CoMP-like schemes
• Spider (soft) handover schemes
• Using moving base stations to serve both in-vehicle and out-of-vehicle users
• Opportunistically utilize moving nodes as ad hoc base stations forming hybrid networks consisting of network

infrastructure nodes and less controllable nodes to enable cost efficient services in mega cities

• Dept. of Electrical Engineering, Communication Systems

Slide 51

Potential of Moving Relay Enabled Cellular Networks in Dense Urban Scenarios

• X. Tang, X. Xu, T. Svensson, X. Tao, “Coverage Performance of Joint Transmission for Moving Relay Enabled

Cellular Networks in Dense Urban Scenarios”, in IEEE Access, vol. 5, no. , pp. 13001-13009, 2017.

CJT: Bias based joint processing CoMP NC-MRP: Non-coordinated maximum-received power- based association NC-MRB: Non-coordinated Moving Relay-biased association

• Dept. of Electrical Engineering, Communication Systems

Slide 52

Initial Access: Beam Finding/Tracking at mmWave

Beam finding/tracking – the key for enabling low latency mm-wave access!

Source: mmMAGIC WP4 presentation, ETSI workshop, Sophia-Antipolis, Jan 28, 2016

• Dept. of Electrical Engineering, Communication Systems

Slide 53

• H. Guo, B. Makki, T. Svensson, “A Genetic Algorithm-based Beamforming Approach for Delay-constrained Networks”,

IEEE WiOpt’2017, May 2017.

• H. Guo, B. Makki, T. Svensson, “A Comparison of Beam Refinement Algorithms for Millimeter Wave Initial

Access”,PIMRC’2017 workshops. Montreal, Canada, Oct 2017.

Illustrative Initial Access Result

53

• Dept. of Electrical Engineering, Communication Systems

Slide 54

System Model

Base Station

(M transmit antennas)

…

Mm-wave channel

𝛾 antennas for each car

VUE 1 VUE 𝝊 VUE 2

v 𝑂 = 𝜐 ∗ 𝛾 antennas at the receiver side

VUE 3

v v v

Base Station

…

Mm-wave channel

VUE 1 VUE 𝝊 VUE 2 VUE 3

v Non-Collaborative users (NCU) Collaborative Users (CU) (e.g., platooning vehicles)

• Dept. of Electrical Engineering, Communication Systems

Slide 55

Simulation Results for Cooperative/Non- cooperative Beam Finding

• Simulation of CUs and NCUs for

GA and Tabu. M = 32, 𝜐 = 4, N = 8, k = 0.

• Substantial gains with

cooperative users (CU) over non- cooperative users (NCU) with the two considered Genetic Algorithm (GA) and Tabu search algorithms.

• The performance gain with CU
• ver NCU increases with SNR.
• H. Guo, B. Makki, T. Svensson, “Genetic-Algorithm Based Beam Refinement for Initial Access in Millimeter-Wave

Mobile Networks”, Wiley-Hindawi Wireless Communications and Mobile Computing, Special Issue on Recent Advances in 5G Technologies: New Radio Access and Networking. To appear. CU NCU

• Dept. of Electrical Engineering, Communication Systems

Slide 56

Integrated Access and Wireless Backhaul

Source of illustration: EU H2020 5GPPP mmMAGIC

• Dept. of Electrical Engineering, Communication Systems

Slide 57

Multi-Beam Cooperation with Mobility – in Unstructured Beam-centric mmWave Networks

• Fixed Group Cell: AP1 controls 3 fixed group cells of size 3
• Slide Group Cell: In AP2, the group cell serving MT4 is

dynamically reconfigured in order to follow the movement of the user

Left illustration is taken from:

• X. Xu, X. Tao, C. Wu and P. Zhang, "Capacity and Coverage

Analyses for the Generalized Distributed Cellular Architecture- Group Cell," 2006 International Conference on Communications, Circuits and Systems, Guilin, 2006, pp. 847-851.

3 6 9 5 1 2

IPV6 Network

V

17 18 15 16 14 13 11 12 MT1 MT2 MT4

AP 1 AP 2

19 20 8 MT5 MT7

Virtual MIMO

MT6 4 7 7 MT3 Group Cell 1 Group Cell 2 Group Cell 3

• Y. Hong, X. Xu, M. Tao, J. Li, T. Svensson, “Cross-tier

Handover Analyses in Small Cell Networks: A Stochastic Geometry Approach”, ICC’2015

Equal long-term averaged biased DL-RSS coverage Boundaries (ESBs) in a two-tier small cell network with macro cells (blue squares) and small cells (red triangles).

• Dept. of Electrical Engineering, Communication Systems

Slide 58

Hybrid Automatic Repeat request (HARQ)

Current focus: Fast HARQ protocols for delay constrained applications. HARQ

(Massive) point-to- point MIMO Coordinated multiuser/multi-link setups Spectrum sharing networks Finite block-length analysis of HARQ Relay networks Green SISO setups Dense heterogeneous networks CoMP networks

• B. Makki, T. Svensson, G. Caire, M. Zorzi, “Fast HARQ over Finite Blocklength Codes: A Technique for Low-Latency

Reliable Communication”, submitted to IEEE TWireless

• Dept. of Electrical Engineering, Communication Systems

Slide 59

Integrated Moving Networks: Mutual Opportunities

• Mutual benefits!
• Better mobile systems

efficiency: Vehicles collect side information to improve the resource allocation and performance of the mobile network

• More reliable V2X links:

Connect non-vehicular users to the Traffic Safety/Traffic Efficiency protocols (Pedestrians, cyclists, pets, …)

• New disruptive business
• pportunities: exploiting

vehicle sensed data New opportunities

• Dept. of Electrical Engineering, Communication Systems

Slide 60

Ongoing Research on Integrated Moving Networks at Chalmers

• Enabling technologies for Integrating ad-hoc network elements like moving base

stations to 5G HetNets

– Evolved predictor antenna systems (sensitivity analyses, mm-wave scenarios, …) – Advanced closed loop (massive) MIMO cooperative moving backhaul links based on predictor antenna concept – Design closed-loop and cooperative interference coordination in ultra-dense heterogeneous networks – Pro-active resource allocation utilizing side-information (like road infrastructure information, driving route information, positioning and social networks) for moving base stations – Jointly optimize overall resource allocation to optimally serve in-vehicle and out-vehicle users using hybrid networking consisting of infrastructure and ad-hoc network elements

• Analyze performance of hybrid networks using stochastic geometry (both user

specific and network specific metrics like outage, throughput, latency and energy efficiency)

• Analyze and develop handover schemes in hybrid networks
• Multi-node multi-beam tracking at mm-wave
• Integrated security and communications for automotive 5G scenarios

The work has considered <=6 GHz, current main focus on >6 GHz carrier frequencies

• Dept. of Electrical Engineering, Communication Systems

Slide 61

Take Home Messages

• Cellular-Assisted V2X can do much more than ad-hoc V2X networks
• Vehicles should play an active role in such networks
• Advanced infrastructure links would be a key enabler
• CSIT based closed loop transmission enabled by Predictor antennas enables

– Enhanced robustness and energy efficiency in the moving backhaul link – Potential spatial multiplexing in the moving backhaul link – Potential to fully integrate moving small cells in a HetNets concept

• CoMP-like interference coordination
• Efficient soft (spider) handover approaches
• Using moving BSs to serve outdoor users also in interference limited scenarios
• Additional opportunities to explore

– Full duplex in the moving backhaul links – mm-wave communication in MNs – Context information in MNs for mutual benefit of VUEs and UEs – Integrated security and communications for automotive 5G scenarios

• Integrated Moving Networks: Mutual opportunities for enhanced mobile

networks and ITS services!

Further reading on MNs: A. Osseiran, J. Monserrat, O. Queseth, P. Marsch, …, T. Svensson, et, al. ”5G Mobile Communications Technology ”, Cambridge University Press, June 2016. ISBN: 9781107130098. – Chapter 11

• Dept. of Electrical Engineering, Communication Systems

Slide 62

Relevant Recent Publications

• H. Guo, B. Makki, T. Svensson, “Genetic-Algorithm Based Beam Refinement for Initial Access in Millimeter-Wave Mobile Networks”,

Wiley-Hindawi Wireless Communications and Mobile Computing, Special Issue on Recent Advances in 5G Technologies: New Radio Access and Networking. To appear.

• H. Guo, B. Makki, T. Svensson, “A Comparison of Beam Refinement Algorithms for Millimeter Wave Initial Access”,PIMRC’2017

workshops.

• H. Guo, B. Makki, T. Svensson, “A Genetic Algorithm-based Beamforming Approach for Delay-constrained Networks”, IEEE WiOpt’2017,

May 2017.

• X. Tang, X. Xu, T. Svensson, X. Tao, “Coverage Performance of Joint Transmission for Moving Relay Enabled Cellular Networks”, in IEEE

Access, vol. 5, no. , pp. 13001-13009, 2017.

• D. Phan-Huy, M. Sternad, T. Svensson, W. Zirwas, B. Villeforceix, F. Karim, Salah El Ayoubi, “5G on board: how many antennas do we

need on connected cars?”, IEEE Globecom’2016 Workshop on 5G RAN Design

• A. Osseiran, J. Monserrat, O. Queseth, P. Marsch, …, T. Svensson, et al. ”5G Mobile Communications Technology”, Cambridge University

Press, June 2016. ISBN: 9781107130098. [Chapter 11 Interference Management, Mobility Management and Dynamic Reconfiguration]

• Y. Sui, I. Guvenc, T. Svensson “Interference Management for Moving Networks in Ultra-Dense Urban Scenarios,” EURASIP Journal on

Wireless Communications and Networking, 2015. (2015) 2015:111.

• D. Phan-Huy, M. Sternad, T. Svensson, "Making 5G Adaptive Antennas Work for Very Fast Moving Vehicles," Intelligent Transportation

Systems Magazine, IEEE , vol.7, no.2, pp.71,84, Summer 2015.

• J. Li, E. Björnson, T. Svensson, T. Eriksson, M. Debbah, “Joint Precoding and Load Balancing Optimization for Energy-Efficient

Heterogeneous Networks”, IEEE Transactions on Wireless Communications, vol.14, no.10, pp.5810-5822, Oct. 2015.

• J. Li, E. Björnson, T. Svensson, T. Eriksson, and M. Debbah “Optimal Design of Energy-Efficient HetNets: Joint Precoding and Load

Balancing”, IEEE International Conference on Communications, ICC’2015, London, UK, 2015. Best paper award.

• N. Jamaly, R. Apelfrojd, A.B. Martinez, M. Grieger, T. Svensson, M. Sternad, G. Fettweis, “Analysis and Measurement of Multiple

Antenna Systems for Fading Channel Prediction in Moving Relays”, IEEE European Conference on Antennas and Propagation (EuCAP’2014), The Hague, The Netherlands, April 2014.

• Y. Sui, I. Guvenc, and T. Svensson, “On the Deployment of Moving Networks in Ultra-dense Urban Scenarios”, 1st International

Conference on 5G for Ubiquitous Connectivity (5GU), 2014, Invited paper.

• Y. Sui, A. Papadogiannis, J. Vihriälä, M. Sternad, W. Yang, T. Svensson, “Moving Cells: A promising solution to boost performance for

vehicular users”, Special Issue IEEE Communications Magazine, 51, (6), 2013.

• D.T. Phan-Huy, M. Sternad, T. Svensson, ”Adaptive large MISO downlink with predictor antenna array for very fast moving vehicles”,

IEEE ICCVE’2013, Dec 2013, Las Vegas.

• Dept. of Electrical Engineering, Communication Systems

Slide 63

Relevant Recent Publications cont.

• Y. Sui, T. Svensson, “Uplink Enhancement of Vehicular Users by Using D2D Communications,” IEEE Workshop on

D2D Communication, GLOBECOM 2013, Atlanta.

• Y. Sui, Z. Ren, W. Sun, T. Svensson, P. Fertl, “Performance Study of Fixed and Moving Relays for Vehicular Users

with Multi-cell Handover under Co-channel Interference”, IEEE ICCVE’2013, Dec 2013, Las Vegas. Best paper award.

• Y. Sui, A. Papadogiannis, W. Yang, T. Svensson, “The Energy Efficiency Potential of Moving and Fixed Relays for

Vehicular Users”, IEEE Vehicular Technology Conference (VTC Fall), 2013, Las Vegas.

• Y. Sui, A. Papadogiannis, W. Yang, T. Svensson, ”Performance Comparison of Fixed and Moving Relays under Co-

channel Interference”, IEEE 4th Int. Workshop on Heterogeneous and Small Cell Networks (HetSNets), GLOBECOM 2012.

• M. Sternad, M. Grieger, R. Abildgaard-Olesen, T. Svensson, D. Aronsson, A. M. Belen, “Using “Predictor

Antennas” for Long-Range Prediction of Fast Fading for Moving Relays”, IEEE Wireless Communications and Networking Conference (WCNC), 2012.

• A. Papadogiannis, Y. Sui, T. Svensson, ”The Potential of a Hybrid Fixed/User Relay Architecture -- A Performance

Analysis”, IEEE Vehicular Technology Conference, VTC 2012-Fall, Quebec City, Canada, Sep. 2012. ISBN/ISSN: 978-1-4673-1880-8.

• A. Papadogiannis, M. Färber, A. Saadani, M. D. Nisar, P. Weitkemper, Y. Sui, T. Svensson, D. Ktenas, N. Cassiau, T.

de Moraes, Thiago-Martins, “Advanced Relaying Concepts for Future Wireless Networks”, Future Network and Mobile Summit (FUNEMS 2012), Berlin, Germany, July 2012.

• Y. Sui, A. Papadogiannis, T. Svensson, “The Potential of Moving Relays - A Performance Analysis”, 2012 IEEE

75th Vehicular Technology Conference: VTC2012-Spring 6-9 May 2012, Yokohama, Japan.

• Dept. of Electrical Engineering, Communication Systems

Slide 64

7th Globecom’2017 Workshop on International Workshop on Emerging Technologies for 5G and Beyond Wireless and Mobile Networks (ET5GB)

Sun Dec 9, 2018, Abu Dhabi

Main topics:

• Novel radio access network (RAN) architectures
• Novel enablers for wireless networks
• Advanced radio resource management (RRM) techniques
• Emerging technologies in physical layer
• Verticals and novel services
• mm-wave and THz communications
• Energy efficiency
• Spectrum
• Prototype and test-bed for 5G+ technologies

Workshop Chairs:

• Tommy Svensson, Chalmers U. of Technology, Sweden
• Halim Yanikomeroglu, Carleton University, Canada
• Peiying Zhu, Huawei Technologies, Canada

Technical Program Chairs:

• Huseyin Arslan, University of South Florida, Tampa, USA
• Lingjia Liu, University of Kansas, USA
• Charlie (Jianzhong) Zhang, Samsung Electronics, USA

http://www.et5gb.com/ http://www.ieee-globecom.org/

• Keynote Talks
• Posters
• Panel

Welcome to join the vivid discussions!

• Dept. of Electrical Engineering, Communication Systems

Slide 65

Extra slides

• Dept. of Electrical Engineering, Communication Systems

Slide 66

Moving Base Station (MBS) to form Moving Cells/Networks

Extra slides for

• Dept. of Electrical Engineering, Communication Systems

Slide 67

ABS Configuration for Micro BSs

• Dept. of Electrical Engineering, Communication Systems

Slide 68

Simulation Assumptions (1/3)

Component Configuration Parameters Macro BS Height: 5 meters above the top of the middle building Maximum transmit power (per 10 MHz): 43 dBm Carrier: 800 MHz Bandwidth: 20 MHz Antenna configuration: 17 dBi, 3 sectors (one antenna per sector), 0º, 120º and 240º with respect to the north Micro BS Height: 10 meters above the ground close to middle point of south and east walls Position: see fig. 1 Maximum Transmit power (per 10 MHz): 30 dBm Carrier: 2.6 GHz Bandwidth: 80 MHz Cell range expansion bias: 5 dB Antenna configuration: 17 dBi gain, 2 sectors (one antenna per sector), pointing to the main street with an angle of 20º with respect to the closest wall

• Dept. of Electrical Engineering, Communication Systems

Slide 69

Simulation Assumptions (2/3)

Component Configuration Parameters Outdoor UEs (Macro UE, Micro UE) Speed: 0 to 3 km/h Height: 1.5 meters above the ground Position: uniformly randomly distributed, 50 UEs per road. Cell selection: based on received power with 5 dB cell range expansion (CRE) bias for micro cells. Receiver noise figure: 9 dB Buildings and Streets 9 buildings 120 meters by 120 meters with 6 floors (3.5 meter height of each floor) Road width 21 meters (including sidewalks and parking lanes)

• Macro cells
• Modified Proportional Fairness scheduling. MNs are regarded as super users that aggregate

the traffic of the VUEs they are serving

• Micro cells
• Regular Proportional Fairness scheduling based on throughput

• Dept. of Electrical Engineering, Communication Systems

Slide 70

Component Configuration Parameters Moving Network Full-duplex Speed: 50 km/h Height: 3.5 meters above the ground Position: randomly generated according to a Poisson distribution with λ= 0.5 in each direction of the road. Maximum transmit power (per 10 MHz): 10 dBm Carrier: 800MHz for backhaul links, and 2.6 GHz for access links Bandwidth: 20 MHz for backhaul links, and 80 MHz for access links Antenna configuration: single antenna, 0 dBi omnidirectional antenna Receiver noise figure: 5 dB VUEs Height: 1.5 m above the ground Position: uniformly randomly distributed inside a vehicle. Number of VUEs in each vehicle: 1). uniformly from the interval [1, 50]; 2). 25 VUEs per vehicle. Cell selection: always connect to the MN of their own vehicles Receiver noise figure: 9 dB

Simulation Assumptions (3/3)

• Dept. of Electrical Engineering, Communication Systems

Slide 71

Illustration of Throughput of MNs without Inter Cell Interference Cancellation (ICIC) in Micro Cells

MN backhaul using Maximum Ratio Combining (MRC) with 4 antennas. VPL is 15dB. cdf

Interference management is important for both the backhaul links and access links!