Accessing from the Sky: UAV Communications for 5G and Beyond€¦ · Y. Zeng, Q. Wu, and R. Zhang, “Accessing from the Sky: a tutorial on UAV communications for 5G and beyond,’’

Accessing from the Sky: UAV Communications for 5G and Beyond

Rui Zhang National University of Singapore

(e-mail: [email protected])

Yong ZengThe University of Sydney, Australia

(e-mail: [email protected])

IEEE ICC20 May 2019, Shanghai, China

ICC, 2019 1

Outline

Introduction (Rui Zhang)

UAV Communication Fundamentals (Rui Zhang)

UAV-Assisted Wireless Communications (Yong Zeng)

Cellular-Connected UAV (Yong Zeng)

Extensions and Future Work (Yong Zeng)

ICC, 2019 2

Outline Introduction

UAV communication requirement Wireless technologies for UAV The new paradigm: integrating UAVs into cellular

• Cellular-connected UAV• UAV-assisted wireless communications

What’s New?

UAV Communication Fundamentals

UAV-Assisted Wireless Communications

Cellular-Connected UAV

Extensions and Future Work ICC, 2019 3

Rui Zhang, National University of SingaporeIntroduction

UAV: Whose Time is Coming

Create more than 100,000 new jobs in US over the next 10 years Numerous applications: military, traffic control, cargo delivery, precision

agriculture, video streaming, aerial inspection, rescue and search, ……

Source: SDI analysis. https://www.marketresearch.com/product/sample-8691316.pdf

4ICC, 2019

Rui Zhang, National University of Singapore

Aerial photography

5

Drone Delivery

Aerial Inspection Precision Agriculture

ICC, 2019

Introduction


Traffic Monitoring

6ICC, 2019

Introduction

H. Menouar, et. al, “UAV-Enabled Intelligent Transportation Systems for the Smart City: Applications and Challenges”, IEEE Commun. Mag., Mar. 2017.


Airborne Communication Platform

Y. Zeng, R. Zhang, and T. J. Lim, “Wireless communications with unmanned aerial vehicles: opportunities and challenges,” IEEE Commun. Mag., May 2016.

7ICC, 2019

Introduction


FAA Rules for Small UAS June 2016: US Federal Aviation Administration (FAA) released new

rules for commercial use of small unmanned aircraft systems (UAS) (Part 107) Small UAS: 55 pounds (25 kg) Visual line-of-sight (VLoS) only Daylight-only operation Maximum ground speed: 100mph (161km/h) Maximum altitude: 400 feet (122 m) Operations are allowed in uncontrolled airspace (Class G) without air

traffic control (ATC) permission The road ahead:

Beyond VLoS operation? Completely autonomous flying?

8

Source: FAA, “Operation and Certification of Small Unmanned Aircraft Systems”, https://www.faa.gov/uas/media/RIN_2120-AJ60_Clean_Signed.pdf

ICC, 2019

Introduction


Wireless Communications for UAVs: Basic Requirement Control and Non-Payload

Communications (CNPC) Ensure safe, reliable, and effective

flight operation Low data rate, high reliability, high

security, low latency Telemetry (UAV status reporting) Command and control Navigation aids Sense and avoid (S&A) Air traffic control (ATC) relay,……

Payload Communications Application specific information Typically higher rate than CNPC CNPC information flows [ITUReportM.2171]

Control Station UAV

9ICC, 2019

Introduction

ITU, “Characteristics of unmanned aircraft systems and spectrum requirements to support their safe operation in non-segregated airspace,” Tech. Rep. M.2171, Dec. 2009.


3GPP UAV Communication Requirement

10ICC, 2019

Introduction

Data Type Data Rate Reliability Latency

DL (BS to UAV)

Command and control

60-100 Kbps

10−3 packet error rate

50 ms

UL (UAV toBS)

Command and control

60-100 Kbps

10−3 packet error rate

--

Application data

Up to 50 Mbps

-- Similar toterrestrial user

3GPP TR 36.777: “Technical specification group radio access network: study on enhanced LTE support for aerial vehicles”, Dec. 2017.


China Mobile Requirement for Typical Payload

11ICC, 2019

Introduction

UAV Application Height coverage Payload traffic latency

Payload data rate (DL/UL)

Drone delivery 100 m 500 ms 300 Kbps/200 Kbps

Drone filming 100 m 500 ms 300 Kbps/30 Mbps

Access point 500 m 500 ms 50 Mbps/50 Mbps

Infrastructureinspection 100 m 3000 ms 300 Kbps/10 Mbps

Drone fleet show 200 m 100 ms 200 Kbps/200 Kbps

Precisionagriculture 300 m 500 ms 300 Kbps/200 Kbps

“China mobile technical report: Internet of drones (in Chinese),”http://www.jintiankansha.me/t/AE9FsWW9tc


Wireless Technologies for UAV Communications (1/4) Direct ground-to-UAV communications

Simple, low cost Unlicensed spectrum (e.g., 2.4 GHz)

Main limitations: Unreliable, insecure Vulnerable to interference Limited data rate Visual line of sight (LoS) operation Difficult to legitimately monitor and manage

12ICC, 2019

Introduction


Wireless Technologies for UAV Communications (2/4) Satellite-UAV communications

Communication and Internet access via satellites Global coverage

Main limitations: Costly Heavy/bulky/energy-consuming communication equipment High latency Large signal attenuation

13ICC, 2019

Introduction


Wireless Technologies for UAV Communications (3/4) Flying ad-hoc network

Dynamically self-organizing and infrastructure-free Robust and adaptable Support for high mobility

Main limitations: Low spectrum efficiency Intermittent connectivity Complex routing protocol

14ICC, 2019

Introduction


Wireless Technologies for UAV Communications (4/4) Cellular network for UAV

UAV communications enabled by cellular infrastructure and technologies Almost ubiquitous accessibility worldwide Cost-effective, superior performance and scalability

Main limitations: Unavailable in remote areas Potential interference with terrestrial communications

15ICC, 2019

Introduction


Integrated Network for UAV Communications

16ICC, 2019

Introduction

Y. Zeng, Q. Wu, and R. Zhang, “Accessing from the Sky: a tutorial on UAV communications for 5G and beyond,’’ submitted to Proceedings of the IEEE, Invited Paper, available on arXiv


Integrating UAVs into Cellular: A Win-Win Technology Integrating UAVs into cellular networks (LTE, 5G, B5G…) For UAV industry:

More reliable and secure CNPC Unlimited operation range without relying on satellite High capacity enabled by advanced cellular techniques Cost effective: avoid building new infrastructure for UAV

communication Enhanced air traffic management: legitimate control takeover when

needed For cellular industry:

New business opportunities by incorporating aerial users More robust and cost-effective cellular network with aerial

communication platforms

17ICC, 2019

Integrating UAVs into 5G and Beyond

Y. Zeng, J. Lyu, and R. Zhang, “Cellular-connected UAV: potentials, challenges and promising technologies, IEEE Wireless Commun., Feb. 2019.


Integrating UAVs into Cellular: Two Frameworks Cellular-Connected UAV: UAVs as aerial users with their own missions

UAV-Assisted Wireless Communication: UAVs as aerial communication platforms (e.g. BS, relay, access point)

Cellular-Connected UAV UAV-Assisted Wireless Communication

18ICC, 2019




UAV Communications: What’s New?

19

High altitude Up to 300 meters (as user in 3GPP) or even higher (as BS/relay)

High 3D mobility Partially or fully controllable placement/trajectory

New air-to-ground channel characteristic High probability of radio line-of-sight (LoS)

Size, weight and power constraint (SWAP) Limited UAV on-board energy/endurance

New opportunities and challenges for both cellular-connected UAVs and UAV-assisted wireless communications

ICC, 2019


Rui Zhang, National University of SingaporeCellular-Connected UAV

Cellular-Connected UAVs High altitude

3D coverage is challenging: existing BS antenna tilted downwards High 3D mobility

Partially controllable trajectory due to mission requirement (e.g., cargo delivery) Asymmetric downlink/uplink: ultra-reliable CNPC versus high-rate payload data Strong air-to-ground LoS channel

Pro: High degrees of macro-diversity (beneficial for BS association, CoMP) Con: Severe aerial-ground interference (in both uplink and downlink)

Mainly served by antenna side-lobe with current LTE BS

20ICC, 2019


Severe Aerial-Ground Interference

Strong LoS Interference between UAV and terrestrial users Uplink: more interference imposed to neighbouring BSs Downlink: more interference received from neighbouring BSs

21ICC, 2019


Reshaping Cellular Networks to Cover the Sky New techniques needed to enable highly heterogonous network with both

terrestrial and aerial users (more details given later) Possible spectrum allocation:

Protected aviation spectrum licensed to cellular operators for UAV CNPC Cellular spectrum shared by aerial and terrestrial users for payload data

Interference management techniques: Sub-sector in elevation domain 3D beamforming Coordinated resource allocation/CoMP Terrestrial-aerial NOMA Joint trajectory design and communication resource allocation

22ICC, 2019



Cellular-Connected UAVs: Industry Efforts

February 2016 Intel and AT&T Demonstrated the world’s first LTE-connected drone at the 2016 Mobile World Congress

August 2016 Ericsson and China Mobile

Claimed the world’s first 5G-enabled drone prototype field trial in WuXi of China

September 2016 Qualcomm Feasibility proof of drone operation over commercial LTE networks up to 400 feet

March 2017 3GPP Study item on “study on enhanced support for aerial vehicles using LTE”

May 2017 Qualcomm Trial report “LTE Unmanned Aircraft Systems”

November 2017 Huawei Wireless X Labs

Announced the "Digital Sky Initiative“ to “spur development of drone applications and enable the low airspace digitized economy via enhanced low airspace network coverage”

December 2017 3GPP 3GPP TR 36.777, Technical Report (Release 15)

23ICC, 2019

Advantages of UAV-assisted communication On demand deployment, fast response Cost effective Fully controllable mobility: dynamic 3D placement and

movement Short-distance LoS link

Three typical use cases UAV-aided ubiquitous coverage UAV-aided mobile relaying UAV-aided information dissemination/data collection


UAV-Assisted Wireless CommunicationsUAV-Assisted Wireless Communication

24ICC, 2019

UAV-Aided Ubiquitous Coverage Provide seamless coverage within the serving area Application scenarios: fast service recovery after natural disaster BS offloading at hotspot

Rui Zhang, National University of SingaporeUAV-Assisted Wireless Communication

Y. Zeng, R. Zhang, and T. J. Lim, "Wireless communications with unmanned aerial vehicles: opportunities and challenges," IEEE Commun. Mag., May 2016

25ICC, 2019

UAV-Aided Mobile Relaying Connecting two or more distant users or user groups Application scenarios: Emergency response, e.g., between frontline and headquarter Military network



26ICC, 2019

UAV-Aided Information Dissemination/Data Collection Application scenarios:

Periodic sensing IoT communications Multicasting



27ICC, 2019

UAV-Enabled Radio Access Network (RAN): A General Model


J. Zhang, Y. Zeng, and R. Zhang, “UAV-enabled radio access network: multi-mode communication and trajectory design,” IEEE Trans. Signal Process., Oct. 2018.

28ICC, 2019


UAV-mounted Aerial BS: Practical Realizations Tethered UAVs: UAVs connected with ground platform by cables (for power and/or

data backhaul) Examples: AT&T’s “flying cell-on-wings (COWs)”, Facebook’s “Tether-Tenna”, EE’s “Air

Masts” Untethered UAVs: Wireless, free-fly drones Examples: Nokia’s flying-cell (F-Cell) and Facebook’s Aquila

AT&T’s flying COW Nokia’s F-Cell

29

Cable

Source: http://www.zdnet.com/article/at-t-completes-its-first-flying-cow-test-flight/https://www.fiercewireless.com/tech/nokia-bell-labs-uses-drone-to-deliver-breakthrough-f-cell-to-office-rooftop

UAV-Assisted Wireless Communication

ICC, 2019


UAV Communication Fundamentals Channel model Antenna model UAV energy consumption model Performance metric Mathematical formulation



Extensions and Future Work

ICC, 2019 30

UAV Channel Modelling: An OverviewRui Zhang, National University of Singapore

31

UAV Channel Model

ICC, 2019

UAV-UAV UAV-GroundBS-UAV UAV-terminal

Large-scale: PL + shadowing

Free-space path loss model

Customized models for UAV-ground links

Small-scale: multipath fading

Usually negligible

Rayleigh, Rician, Nakagami-m

Rayleigh, Rician, Nakagami-m

Generic narrowband (frequency-flat) fading channel model

Large-scale gain Small-scale fading

Customized Models for UAV-Ground Channels Rui Zhang, National University of Singapore

Customized models to reflect different propagation environment with varying UAV altitude/locations

Three typical categories: Free-space path loss model Altitude/angle-dependent channel parameters

• Altitude/height-dependent path loss exponent, Rician factor…• Elevation angle-dependent path loss exponent, Rician factor…• Depression angle-dependent excess path loss

Probabilistic LoS channel model • Elevation angle-dependent probabilistic LoS model• 3GPP BS-UAV model

Other models

32

UAV Channel Model

ICC, 2019

Free-Space Path Loss ModelRui Zhang, National University of Singapore

Channel power inversely proportional to distance square

No shadowing or small-scale fading Proposed scenarios: BS-UAV and UAV-terminal channels in rural area

and/or with sufficiently high UAV altitude Pros:

Simple, channel highly predictable Useful for offline UAV trajectory design

Cons: Oversimplified in urban environment Fails to model change of propagation environment at different UAV

altitude/locations

33

UAV Channel Model

ICC, 2019

Altitude/Height-Dependent Channel ParametersRui Zhang, National University of Singapore

Channel modelling parameters are functions of UAV altitude Altitude-dependent path loss exponent, shadowing variance, Rician

factor,…

Proposed application scenarios: BS-UAV channels in urban/suburban area Pros:

Captures environmental variations as altitude changes Useful for theoretical performance analysis

Cons: Fails to model the change of propagation condition with horizontal fly

34

UAV Channel Model

ICC, 2019

R. Amorim et al., “Radio channel modeling for UAV communication over cellular networks,” IEEE Wireless Commun. Lett., Aug. 2017.

Elevation Angle-Dependent Channel ParametersRui Zhang, National University of Singapore

Channel modelling parameters are functions of elevation angle Elevation angle-dependent path loss exponent, Rician factor,… Proposed scenarios: BS-terminal channels in urban/suburban area Pros:

Captures environmental variations with 3D flying Useful for theoretical performance analysis

35

UAV Channel Model

ICC, 2019

Cons: Further experimental

verification required

M. M. Azari, F. Rosas, K.-C. Chen, and S. Pollin, “Ultra reliable UAV communication using altitude and cooperation diversity,” IEEE Trans. Commun., Jan. 2018.

Depression Angle-Dependent Excess Path Loss ModelRui Zhang, National University of Singapore

Excess path loss on top of conventional terrestrial path loss Excess path loss is a function of depression (elevation) angle

Proposed scenarios: BS-UAV channels in suburban area

36

UAV Channel Model

ICC, 2019

A. Al-Hourani and K. Gomez, “Modeling cellular-to-UAV path-loss for suburban environments,” IEEE Wireless Commun. Lett., pp. 82–85, Feb. 2018.

Pros: Applicable for both positive

and negative angles Cons:

Difficult for performance analysis

Elevation Angle-Dependent Probabilistic LoS ModelRui Zhang, National University of Singapore

Separately model LoS and NLoS propagations LoS probability increases with elevation angle

Proposed scenarios: UAV-GT channel in urban environment with statistical information of building height/distribution

Pros: Useful for theoretical performance analysis Cons:

Models in statistical sense only, may fail in actual environment Needs experimental verification

37

UAV Channel Model

ICC, 2019

A. Al-Hourani, S. Kandeepan, and S. Lardner, “Optimal LAP altitude for maximum coverage,” IEEE Wireless Commun. Lett., Dec. 2014.

Average Channel Power vs AltitudeRui Zhang, National University of Singapore

38

UAV Channel Model

ICC, 2019

Tradeoff: higher altitude, increased LoS probability but longer link distance


3GPP BS-UAV ModelRui Zhang, National University of Singapore

Separately model LoS and NLoS propagations LoS probability and channel modelling parameters are functions of

UAV altitude and horizontal distance

Proposed scenarios: BS-UAV for Urban Macro (UMa), Urban Micro (UMi), and Rural Macro (RMa)

Pros: Comprehensive models for path loss, shadowing and small-scale fading Useful for numerical simulation

Cons: Too complicated for theoretical performance analysis and trajectory optimization

39

UAV Channel Model

ICC, 2019


Other Models and Future WorkRui Zhang, National University of Singapore

Modified log-distance path loss model with a correction for flight direction

3D geometry-based stochastic MIMO UAV channel model Two-ray channel model Tapped delay line model for wide-band UAV communicationMassive MIMO UAV channel modelMillimeter wave UAV channel model

40

UAV Channel Model

ICC, 2019

D. W. Matolak and R. Sun, “Air-ground channel characterization for unmanned aircraftsystems–Part III: the suburban and near-urban environments,” IEEE Trans. Veh. Tech.,Aug. 2017.

Antenna ModelRui Zhang, National University of Singapore

Isotropic Equal transmit/receive power in all directions Spherical radiation pattern in 3D Hypothetical

Directional antenna with fixed radiation pattern Fixed antenna gain given azimuth/elevation angles By designing antenna geometry, e.g., dish, horn, or using Antenna array with fixed phase shifting

Active antenna arrays Active elements with dynamically controllable phase/magnitude Multi-antenna beamforming/precoding and combining MIMO (similar to terrestrial MIMO communications)

41

Antenna Model

ICC, 2019

Cellular BS 3D Directional Antenna ModelRui Zhang, National University of Singapore

Cellular BS antennas are typically down-tilted electrically/mechanically Electrical down-tilt: vertical ULA with fixed phase shifting

42

Antenna Model

ICC, 2019

Two-lobe approximation:

UAV 3D Directional Antenna ModelRui Zhang, National University of Singapore

UAV main beam directly beneath the aircraft

43

Antenna Model

ICC, 2019

Two-lobe approximation:

ψ𝐻𝐻𝑈𝑈

𝑟𝑟

UAV Energy Consumption ModelUAV Energy Model

Limited on-board energy: critical issue in UAV communication, for both UAV as user or BS/relay

UAV energy consumption: Propulsion energy >> Communication energy Empirical and Heuristic Models:

Empirical model based on measurement result Fuel cost modelled by L1 norm of control force Fuel cost proportional to the square of speed

Analytical Model Closed-form model based on well-established results in aircraft literature Power as a function of speed and acceleration: fixed-wing or rotary-wing

44ICC, 2019


UAV Energy Model (Fixed-Wing) Forces on UAV: weight (W), lift (L), drag (D), and thrust (F) Fixed-wing propulsion power (Straight and level flight)

Parasite power: overcome parasite drag caused by the moving of aircraft Induced power: overcome induced drag resulted from the lift force

45

Y. Zeng and R. Zhang, "Energy-Efficient UAV Communication with Trajectory Optimization," IEEE Trans. Wireless Commun., June 2017.

UAV Energy Model

ICC, 2019


UAV Energy Model (Rotary-Wing) Rotary-wing propulsion power (Straight and level flight)

46

Y. Zeng, J. Xu, and R. Zhang, “Energy minimization for wireless communication with rotary-wingUAV,” IEEE Trans. Wireless Commun., Apr. 2019.

UAV Energy Model

ICC, 2019


Maximum-Endurance and Maximum-Range Speed Maximum-endurance (ME) speed

Maximum-range (MR) speed

47

Energy per unit distance

UAV Energy Model

ICC, 2019


Energy Model Comparison: Fixed-Wing vs. Rotary-WingUAV Energy Model

ICC, 2019 48

Fixed-Wing Rotary-Wing

Convexity with respect to 𝑉𝑉 Convex Non-convex

Components Induced and parasite Induced, parasite, and blade profile

𝑉𝑉 = 0 Infinity Finite


Energy Model with General Level Flight (Fixed-Wing)UAV Energy Model

Change in kinetic energyWork required to overcome air resistance

Only depends on speed and centrifugal acceleration (causing heading change)

Independent of actual location or tangential acceleration (causing speed change)

Work-energy principle interpretation

ICC, 2019 49

𝒂𝒂(𝑡𝑡)

𝒗𝒗(𝑡𝑡)

𝒂𝒂⊥

(𝑡𝑡)

𝒂𝒂||(𝑡𝑡)

𝒂𝒂⊥𝟐𝟐 (𝑡𝑡)


UAV Communication: Performance MetricPerformance Metric

Signal to interference-plus-noise ratio (SINR) Outage/coverage probability Communication throughput/delay Spectral/energy efficiency All dependent on UAV location/trajectory

ICC, 2019 50


Performance Metric: Expected Communication Rate

ICC, 2019 51

Ground node location: 𝒘𝒘𝑘𝑘 ∈ 𝑅𝑅3, UAV trajectory: 𝒒𝒒 𝑡𝑡 ∈ 𝑅𝑅3, 0 ≤ 𝑡𝑡 ≤ 𝑇𝑇

UAV-GN distance: 𝑑𝑑𝑘𝑘 𝑡𝑡 = 𝒒𝒒 𝑡𝑡 − 𝒘𝒘𝑘𝑘 , elevation angle: 𝜃𝜃𝑘𝑘 𝑡𝑡 = sin−1 𝐻𝐻𝑢𝑢(𝑡𝑡)𝑑𝑑𝑘𝑘(𝑡𝑡)

Channel model: Large-scale channel power gain model:

Elevation angle-dependent LoS probability

Two levels of randomness: random LoS link and random small-scale fading Expected channel power gain:

q(t): trajectory

wk

HU(t)𝜃𝜃𝑘𝑘(𝑡𝑡)

dk(t)

Performance Metric Rui Zhang, National University of Singapore

ICC, 2019 52

Communication rate: Average aggregated communication rate:

Trajectory affects both link distance and LoS probability Rate approximation:

Approximation accuracy depends on propagation environment. It becomes exact for free-space channel model.

Performance Metric: Expected Communication Rate

q(t): trajectory

wk

HU(t)𝜃𝜃𝑘𝑘(𝑡𝑡)

dk(t)

Performance Metric Rui Zhang, National University of Singapore

Joint Trajectory-Communication Optimization: Generic Formulation

ICC, 2019 53

𝒒𝒒(𝑡𝑡): trajectory

𝒓𝒓(𝑡𝑡): commun. resource

U: utility functions, e.g., communication rate, SINR, coverage probability, spectrum/energy efficiency

fi: trajectory constraints, e.g., speed constraint, obstacle/collision avoidance gi: communication resource constraints, e.g., power, bandwidth hi: coupled constraints, e.g., interference temperature, minimum SINR

requirement

Mathematical Formulation Rui Zhang, National University of Singapore

Practical Constraints on UAV Trajectory

ICC, 2019 54

Maximum/minimum altitude:

Initial/final status:

Maximum/minimum speed:

Maximum acceleration:

Obstacle avoidance:

Collision avoidance:

No-fly zone (cubic):


Fundamental Tradeoffs in UAV Trajectory and Communication Design

55

Throughput-Delay Tradeoff Throughput-Energy Tradeoff Delay-Energy Tradeoff

Q. Wu, L. Liu, and R. Zhang, “Fundamental tradeoffs in communication and trajectory design for UAV-enabled wireless network,” IEEE Wireless Communications, Feb. 2019.

ICC, 2019


Conclusions of Part I Rui Zhang, National University of Singapore

56

Integrating UAVs into 5G and beyond: a new and promising paradigm to embrace the forthcoming IoD (internet-of-drones) era

Cellular-Connected UAV: UAV as aerial user

UAV-Assisted Wireless Communication: UAV as BS/relay/AP

UAV Communication Fundamentals Channel/antenna/energy models Performance metrics and mathematical formulation for

communication and trajectory co-design Fundamental throughput-delay-energy tradeoffs

More details will be given next and much more to be done …ICC, 2019



UAV-Assisted Wireless Communications Performance analysis UAV placement Joint communication and trajectory optimization Energy-efficient UAV communication



ICC, 2019 57

UAV-Assisted Communications: Basic Models

ICC, 2019 58

Downlink

v

u1

s1

v1v2

s2

u2

Multi-UAV Interference Channel

Interference

S1 D1

s1s1

Relaying

u1 u2

s

v

Multicasting

u1 u2

s1s2

v

u1 u2

s1 s2

Uplink

v

UAV-Assisted Communications Yong Zeng, The University of Sydney

Performance analysis Static UAV platform

• Deterministic modelling of UAV location• Stochastic modelling of UAV location: 2D/3D Poisson Point Process; Binomial

Point Process for small-size UAV network Flying UAV platform

• Deterministic modelling of UAV trajectory: circular/straight• Stochastic modelling of UAV trajectory: stochastic trajectory process

UAV placement No user location information (ULI): maximize covered area Perfect ULI: maximize number of covered users, communication throughput,

minimize number of UAVs given per-user requirement Partial ULI

Trajectory and communication co-design New design DoF by trajectory optimization

Energy-efficient UAV communication

Recent Results: An Overview UAV-Assisted Wireless Communications

59ICC, 2019

Yong Zeng, The University of Sydney

UAV Placement: Spiral-Based AlgorithmUAV Placement

60

Minimize required number of UAVs to ensure all terminals are covered

Core-sets based algorithm, optimal, but with exponential complexity

Proposed spiral-based BS placement algorithm

Example with 80 terminals: Proposed spiral: 11 BSs Optimal core-sets: 11 BSs Benchmark strip-based: 13

BSs

J. Lyu, Y. Zeng, R. Zhang, and T. J. Lim, "Placement Optimization of UAV-Mounted Mobile Base Stations", IEEE Commun. Letters, Mar. 2017.

ICC, 2019


Optimal UAV Altitude with Adjustable BeamwidthUAV Placement

61

UAV antenna with adjustable beamwidth

Fly-hover-communicate protocol Downlink multicasting: altitude as high as

possible, optimal beamwidth decreases with altitude

Downlink broadcasting: altitude as low as possible, beamwidth as small as possible

Uplink multiple access: altitude/beamwidthdoesn’t matter

H. He, S. Zhang, Y. Zeng, and R. Zhang, "Joint altitude and beamwidth optimization for UAV-enabled multiuser communications," IEEE Commun. Lett., Feb. 2018.

ICC, 2019


Probabilistic LoS Channel model Large-scale channel power model for LoS and NLoS conditions

LoS probability:

Expected channel gain:

Exploiting UAV Mobility: How Much Can We Gain?

ICC, 2019 62

d2D

HU

V

𝜃𝜃

d

UAV flies towards the ground terminal

Double effects to improve the channel quality: Shorter link distance Less signal obstruction

𝜅𝜅 < 1: additional attenuation for NLoS

𝐸𝐸[𝛽𝛽(𝑑𝑑)]

Trajectory and Communication Co-Design Yong Zeng, The University of Sydney

ICC, 2019 63

0 5 10 15 20 25 30 35 40 45 50

t (s)

-140

-135

-130

-125

-120

-115

-110

-105

-100

-95

Path

loss

(dB)

LoS

NLoS

Exploiting UAV Mobility: How Much Can We Gain?

0 5 10 15 20 25 30 35 40 45 50

t (s)

-140

-135

-130

-125

-120

-115

-110

-105

-100

-95

Aver

age

chan

nel p

ower

gai

n (d

B)

Channel gain for LoS and NLoSLoS probability

Expected channel gain

Initial distance d2D 1000 m

UAV altitude Hu 100 m

Flying speed v 20 m/s

Path loss exponent α 2.3

Reference channel gain β0

-50 dB

Probabilistic LoSmodel parameters

𝑎𝑎 = 10, 𝑏𝑏 = 0.6,𝜅𝜅 = 0.01

40 dB

23 dB


Exploiting UAV Mobility for Communication: Key Points

ICC, 2019 64

Depending on flying distance and propagation environment, significant gain can be achieved by exploiting UAV mobility

The gain is achieved beyond the conventional communication resource allocation

Motivate the exploitation of UAV trajectory optimization as a newdegree of freedom (DoF) for communication performance enhancement

Main philosophy: bring UAV closer to ground terminals Two related classical problems in graph theory:

Travelling Salesman Problem (TSP) Pickup and Delivery Problem (PDP)

Useful techniques for trajectory and communication co-design Block-coordinate descent Successive convex approximation


Path Planning: Travelling Salesman Problem

ICC, 2019 65

Travelling salesman problem (TSP): Given 𝐾𝐾 cities and the distances between each pair of cities, find the shortest route that visits each city and returns to the origin city

NP-hard problem Complexity by exhaustive search: 𝐾𝐾! Can be formulated as an integer linear programming problem Many heuristic algorithms and exact (up to tens of thousands of cities)

solutions have been proposed

u1 u2

s1 s2

Uplink/downlink

v UAV path planning: For UAV-enabled communications with ground users, determine the flying path to serve them sequentially

Intuition: fly to each user as close as possible

s3

u3


Path Planning: Travelling Salesman Problem

ICC, 2019 66


Variations of Travelling Salesman Problem

ICC, 2019 67

The standard TSP requires the traveler return to the origin city For UAV communications, the UAV may not necessarily return to the

location where it starts the mission, and the initial and/or final locations may be pre-specified

TSP Variation 1: No return TSP Variation 2: No return, specified initial and final locations 𝒒𝒒0 and 𝒒𝒒𝐹𝐹 TSP Variation 3: No return, specified initial location 𝒒𝒒0, any final location


Y. Zeng, X. Xu, and R. Zhang, “Trajectory design for completion time minimization in UAV-enabled multicasting”, IEEE Trans. Wireless Commun., April 2018.

TSP Variation 1: No Return

ICC, 2019 68

Add a dummy city whose distances to all the 𝐾𝐾 cities are 0 Solve the TSP problem for the 𝐾𝐾 + 1 cities Remove the two edges associated with the dummy city

00

0

Dummy city


TSP Variation 1: No Return

ICC, 2019 69

Standard TSP No Return


TSP Variation 2: Specified Initial and Final Locations

ICC, 2019 70

Add a dummy city, whose distance to both 𝒒𝒒0 and 𝒒𝒒𝐹𝐹 set to 0 The distance to all other cities set to sufficiently large Solve the TSP for the K+1 cities Remove the two edges associated with the dummy city

0

∞

Dummy city

0

∞


ICC, 2019 71

No Return, any initial/final No Return, specified initial and final

TSP Variation 2: Specified Initial and Final Locations Trajectory and Communication Co-Design Yong Zeng, The University of Sydney

Travelling Salesman Problem with Neighborhood

ICC, 2019 72

When the total operation time 𝑇𝑇 is small, the UAV may not be able to visit all users

TSP with neighborhood (TSPN): Given 𝐾𝐾 cities and the neighborhoods ofeach city, find the shortest route that visits each neighborhood once

A generalization of TSP, also NP-hard

𝒫𝒫: set of all 𝐾𝐾! possible permutations


Travelling Salesman Problem with Neighborhood

ICC, 2019 73

One heuristic solution First, find the visiting order based on TSP by ignoring the neighborhoods Then use convex optimization to find the optimal visiting location within

each neighborhood Update visiting order by TSP over the visiting locations. Repeat.

The neighborhood size needs to be optimized based on communication requirement

TSP TSPN


Pickup-and-Delivery Problem (PDP)

ICC, 2019 74

For UAV-enabled multi-pair relaying, determine the UAV flying path Information-causality constraint: UAV needs to first receive data from a

source before forwarding to its destination Pickup-and-Delivery Problem (PDP): A generalization of TSP with

precedence constraints: for each source-destination pair, visit source before destination.

NP-hard, while algorithms for high-quality solutions exist PDP with neighborhood (PDPN)

S1 S2

s1

s2

Multipair relaying

vs1

D1

D2

s2


UAV Path Planning with TSPN and PDPN

ICC, 2019 75

TSPN PDPN



Limitations of TSP/PDP For Trajectory Optimization

ICC, 2019 76

Suboptimal trajectory in general: Straight flight between waypoints only, while optimal trajectory for

communication may be curved in general Not catering to interference mitigation

Ignores various trajectory constraints: Obstacle avoidance, maximum/minimum speed, no-fly zone….

Only gives the UAV flying path, but trajectory optimization includes both path planning and speed optimization

UAV trajectory is highly coupled with communication resource allocation. Thus, a joint design is more desirable.

A general framework: joint UAV trajectory and communication resource allocation optimization, by using Successive convex approximation Block coordinate descent


Joint Trajectory-Communication Optimization: Generic Formulation

ICC, 2019 77

𝒒𝒒(𝑡𝑡): trajectory

𝒓𝒓(𝑡𝑡): commun. resource

The continuous-time representation of trajectory involves infinite number of variables

Discretization is necessary for optimization and computation purposes Two discretization methods: time discretization and path discretization


Trajectory Discretization

ICC, 2019 78

Time discretization: Time horizon [0,𝑇𝑇] discretized with sufficiently small slot length: 𝑇𝑇 = 𝑁𝑁𝛿𝛿𝑡𝑡 𝒒𝒒 𝑡𝑡 , 0 ≤ 𝑡𝑡 ≤ 𝑇𝑇, 𝒒𝒒 1 ,𝒒𝒒 2 , … …𝒒𝒒 𝑁𝑁 Linear state-space model:

Path discretization: UAV path sampled into finite line segments with small segment length 𝒒𝒒 𝑡𝑡 , 0 ≤ 𝑡𝑡 ≤ 𝑇𝑇, 𝒒𝒒1,𝒒𝒒2, … …𝒒𝒒𝑀𝑀 , 𝑇𝑇1,𝑇𝑇2, … …𝑇𝑇𝑀𝑀 𝑇𝑇𝑚𝑚: duration for the mth segment


Time vs Path Discretization

ICC, 2019 79

Path discretization: generalized time discretization with variable slot length

Time Discretization Path Discretization

Pros • Equal time slot length• Linear state-space representation• Incorporate maximum

acceleration constraint easily

• Fewer variables if UAV hovers or flies slowly

• No need to know T a priori

Cons • Excessively large number of time slots when UAV moves slowly

• Needs to know T a priori

• More variables if UAV flies with high/maximum speed most of the time

0 𝑇𝑇 = 𝑁𝑁𝛿𝛿𝑡𝑡Time discretization: 𝑇𝑇 must known

𝛿𝛿𝑡𝑡 2𝛿𝛿𝑡𝑡

𝒒𝒒[1] 𝒒𝒒[2] 𝒒𝒒[𝑁𝑁]……

Path discretization:𝑇𝑇 can be unknown𝒒𝒒1 ……𝑇𝑇1 𝑇𝑇2

𝒒𝒒2 𝒒𝒒𝑀𝑀


𝑇𝑇 =∑𝑇𝑇𝑚𝑚

Block Coordinate Descent

ICC, 2019 80

Time or path discretization converts the problem into a more tractable form The joint trajectory and resource optimization problems are usually difficult to

be directly solved Block coordinate descent: alternately update one block of variables (say,

trajectory) with the other (resource allocation) fixed. Monotonically converge to a locally optimal solution.

Optimize 𝒒𝒒[𝑛𝑛]Optimize 𝒓𝒓[𝑛𝑛]

𝑙𝑙 = 𝑙𝑙 + 1

𝒒𝒒(𝑙𝑙)[𝑛𝑛] 𝒒𝒒(𝑙𝑙+1)[𝑛𝑛]𝒓𝒓(𝑙𝑙+1)[𝑛𝑛]


Successive Convex Approximation

ICC, 2019 81

Even with given resource allocation, UAV trajectory optimization is usually non-convex, and thus difficult to solve

Non-concave objective functions: e.g., rate maximization Non-convex constraints: e.g., obstacle/collision avoidance, minimum speed Successive convex approximation (SCA):

local optimization via solving a sequence of convex problems converges to a KKT solution if appropriate local bounds are found

• Convex optimization problem• Solution is feasible to the original

non-convex problem

Non-convex optimization problem

Global concave lower bound


Successive Convex Approximation

ICC, 2019 82

Communication rate maximization:

Minimum speed constraint:

Convex optimization based

on lower bounds

Find global concave lower

bounds

𝑙𝑙 = 𝑙𝑙 + 1

𝐴𝐴𝑘𝑘,𝐵𝐵𝑘𝑘: poisitive coefficients depending on 𝒒𝒒(𝑙𝑙)[𝑛𝑛]

𝒒𝒒(𝑙𝑙)[𝑛𝑛] 𝒒𝒒(𝑙𝑙+1)[𝑛𝑛]


ICC, 2019 83

Global Concave Rate Lower BoundTrajectory and Communication Co-Design Yong Zeng, The University of Sydney

UAV Trajectory Optimization (Summary)

ICC, 2019 84

TSP/PDP provide good heuristics for path planning (e.g., initial path) Joint UAV trajectory and communication design is necessary to achieve

optimality Time- and path-discretization are useful techniques to convert to problem

with finite number of variables Block coordinate descent/alternating optimization can be used to update

trajectory and resource allocation iteratively SCA is useful for trajectory optimization sub-problem: the key is to find

global concave lower bounds Block coordinate/SCA converged solution depends on trajectory/path

initialization Other useful techniques Alternating Direction Method of Multipliers (ADMM) Receding horizon/model predictive control Machine learning for online trajectory optimization


Case Studies Example 1: UAV-Enabled Mobile Relaying

Example 2: Multi-UAV Enabled Wireless Network

85ICC, 2019

Y. Zeng, R. Zhang, and T. J. Lim, “Throughput maximization for UAV-enabled mobile relaying systems,” IEEE Trans. Commun., Dec. 2016.

Q. Wu, Y. Zeng, and R. Zhang, “Joint trajectory and communication design for multi-UAV enabled wireless networks,” IEEE Trans. Wireless Commun., Mar. 2018.


System ModelUAV-Enabled Mobile Relaying

ICC, 2019 86

UAV-enabled relay moves at a constant altitude 𝐻𝐻 UAV mobility constraints: (i) Maximum speed; (ii) Initial and final location S-R and R-D channels vary with the UAV location (𝑥𝑥(𝑡𝑡),𝑦𝑦(𝑡𝑡)) Adaptive rate/power transmission by source and relay based on the

predictable channels Problem: maximize throughput via joint power allocation and trajectory

design


87ICC, 2019

Problem FormulationUAV-Enabled Mobile Relaying

UAV mobility constraints:

Channel model: assume line of sight (LoS), perfect Doppler compensation

Information-causality constraints at UAV: only information that has been received from the source can be forwarded to the destination

0 0

: maximum speed: slot index

( , ) : initial location;( , ) : final locationF F

Vnx yx y


88ICC, 2019

Problem FormulationUAV-Enabled Mobile Relaying

aggregate rate at destination

information-causality constraint

power constraint

initial location constraint

speed constraint

final location constraint


89ICC, 2019

Proposed Alternating Power and Trajectory Optimization

UAV-Enabled Mobile Relaying

(P1) is not jointly convex w.r.t. power and relay trajectory Can be approximately solved with alternating optimization Fix trajectory, power allocation is convex Fix power, trajectory optimization is still non-convex, but can be

approximately solved by SCA


90ICC, 2019

Optimal Power Allocation with Fixed TrajectoryUAV-Enabled Mobile Relaying

E.g., UAVs are deployed for surveillance, opportunistic relaying For any fixed trajectory, power allocation is convex

Staircase waterfilling with non-increasing water level at source

Staircase waterfilling with non-decreasing water level at relay


91ICC, 2019

Trajectory Optimization with Fixed PowerUAV-Enabled Mobile Relaying

Successive convex approximation (SCA) based on global concave rate lower bound


92ICC, 2019

Simulation SetupUAV-Enabled Mobile Relaying

Source and destination separated by D=2000 m Maximum UAV speed: 50 m/s Source and relay average transmission power: 10 dBm Simulation scenarios:

Optimized power with fixed trajectory Optimized trajectory with fixed power Jointly optimized power and trajectory


93ICC, 2019

Optimal Power Allocation with Fixed TrajectoryUAV-Enabled Mobile Relaying

Trajectory 1: unidirectionally towards destination Trajectory 2: unidirectionally towards source Trajectory 3: cyclic between source and destination

(a): power allocation at source for trajectory 2

(b): power allocation at relay for trajectory 2

Decreasing water level at source

Increasing water level at relay


94ICC, 2019

Throughput Comparison for Different TrajectoriesUAV-Enabled Mobile Relaying

Mobile relaying significantly outforms static relaying, if UAV trajectory is properly designed

With inappropriate flying path, mobile relaying may even perform worse than static relaying


95ICC, 2019


Trajectories after different iterations of the proposed successive convex optimization algorithm. Large 𝑇𝑇

source node

destination node

final UAV locationinitial UAV location

Optimized Trajectory with Fixed Constant Power AllocationYong Zeng, The University of Sydney

96ICC, 2019


Trajectories after different iterations of the proposed successive convex optimization algorithm. Small 𝑇𝑇

source node

destination node

final UAV locationinitial UAV location

Optimized Trajectory with Fixed Constant Power AllocationYong Zeng, The University of Sydney

97ICC, 2019


Convergence

Fast convergence for the trajectory optimization algorithm


98ICC, 2019


Joint Power and Trajectory Design via Alternating Optimization


Example 2: Multi-UAV Enabled Wireless Network

ICC, 2019 99

Q. Wu, Y. Zeng, and R. Zhang, “Joint trajectory and communication design for multi-UAV enabled wireless networks,” IEEE Trans. Wireless Commun., Mar. 2018.

Yong Zeng, The University of SydneyTrajectory and Communication Co-Design

System ModelMulti-UAV Enabled Wireless Network

ICC, 2019 100

Multi-UAV spectrum sharing, interference channel (IFC)+broadcast channel (BC)

TDMA for user communication scheduling Problem: maximize the minimum average rate via joint user scheduling and

transmit power control as well as UAV trajectory design


Problem FormulationMulti-UAV Enabled Wireless Network

ICC, 2019 101

Minimum rate requirement

UAV mobility constraint

TDMA constraints

power constraint

Initial/final location constraint

collision avoidance constraint


Multi-UAV Enabled Wireless Network

ICC, 2019 102

Proposed Solution

Block coordinate descent algorithm


Proposed Solution: Sub-problem 1Multi-UAV Enabled Wireless Network

ICC, 2019 103

Continuous relaxation on binary variables

Linear program

User Scheduling and Association Optimization


Proposed Solution: Sub-problem 2Multi-UAV Enabled Wireless Network

ICC, 2019 104

Successive convex approximation

UAV Trajectory Optimization


Multi-UAV Enabled Wireless Network

ICC, 2019 105

Proposed Solution: Sub-problem 3

Transmit Power Control

Successive convex approximation


Simulation ResultsMulti-UAV Enabled Wireless Network

ICC, 2019 106

New Interference-mitigation approach: coordinated multi-UAV trajectory design



ICC, 2019 107

Power control under optimized UAV trajectory

Full-power transmission for large inter-UAV distance Small/zero power for small inter-UAV distance



ICC, 2019 108

Throughput-delay tradeoff

Longer flight period achieves higher max-min throughput Longer flight period implies larger user delay on average


Energy-Efficient UAV Communications

Energy-Efficient UAV Communication

ICC, 2019 109


Y. Zeng and R. Zhang, "Energy-Efficient UAV Communication with Trajectory Optimization," IEEE Trans. Wireless Commun., June 2017.

Y. Zeng, J. Xu, and R. Zhang, “Energy minimization for wireless communication with rotary-wingUAV,” IEEE Trans. Wireless Commun., Apr. 2019.

Energy-Efficient UAV CommunicationsEnergy-Efficient UAV Communication

Energy: critical issue for UAV, finite endurance UAV energy consumption:

Communication energy: signal transmission/reception/processing and circuitry (extensively studied in existing literature)

Propulsion energy: maintain UAV aloft and mobile (important but with only limited work, rigorous mathematical model is needed)

UAV energy model Depends on UAV type: fixed-wing vs. rotary-wing A function of UAV velocity and acceleration

Energy-efficient UAV communications: Maximize energy efficiency (bits/Joule) given operation time T Minimize energy expenditure subject to throughput constraint Maximize throughput subject to energy constraint

ICC, 2019 110


Energy Efficiency Maximization: Circular Flight

ICC, 2019 111

Circular trajectory with radius 𝑟𝑟 and speed 𝑉𝑉 UAV energy consumption:

Optimal speed as a function of 𝑟𝑟:

Power decreases with 𝑟𝑟:

Circular flight

Yong Zeng, The University of SydneyEnergy-Efficient UAV Communication

Energy Efficiency Maximization: Circular FlightEnergy-Efficient UAV Communication

Circular flight with radius 𝑟𝑟 and optimal speed 𝑉𝑉 Energy efficiency: number of information bits per unit energy consumption

(bits/Joule)

Smaller 𝑟𝑟: Shorter link distance and higher LoS probability, thus higher communication rate, but more energy consumption

ICC, 2019 112


Energy Efficiency Maximization: General TrajectoryEnergy-Efficient UAV Communication

UAV energy consumption:

Aggregate throughput as a function of UAV trajectory

Energy efficiency in bits/Joule:

ICC, 2019 113



Maximize energy efficiency in bits/Joule via trajectory optimization

ICC, 2019 114

Non-convex, infinite number of variablesMain techniques: time discretization and Successive Convex Approximation (SCA)


Min./Max. speed constraint

Initial/final velocity constraint

Max. acceleration constraint



Maximize energy efficiency in bits/Joule via trajectory optimization

ICC, 2019 115

Non-convex, iteratively solved by Successive Convex Approximation (SCA)


116ICC, 2019

Simulation SetupYong Zeng, The University of Sydney

UAV altitude 100 m

𝑐𝑐1 9.26 × 10−4

𝑐𝑐2 2250Maximum endurance speed

𝑉𝑉em = 30 m/s

Minimum power consumption

100 Watt

Initial location 𝒒𝒒0 = 1000,0 𝑇𝑇

Final location 𝒒𝒒𝐹𝐹 = 0, 1000 𝑇𝑇

Initial/final velocity 𝒗𝒗0 = 𝒗𝒗𝐹𝐹 = 30𝒗𝒗0𝐹𝐹Maximum/minimum speed

𝑉𝑉max = 100 m/s𝑉𝑉min = 3 m/s

Maximum acceleration

𝑎𝑎max = 5 m/s2

Energy-Efficient UAV Communication

Simulation ResultsEnergy-Efficient UAV Communication

ICC, 2019 117



Rate-max trajectory: stay as close as possible with the ground terminal Energy-min trajectory: less acute turning EE-max trajectory: balance the two, “8” shape trajectory

ICC, 2019 118



Rate-max: fly to the ground node as soon as possible and hover around the ground terminal with the minimum required speed

Energy-min trajectory: around the power-minimum speed EE-max trajectory: speed varies based on the distance with ground node

ICC, 2019 119


Energy Minimization with Throughput ConstraintEnergy-Efficient UAV Communication

Minimize UAV energy consumption subject to throughput requirement, via joint trajectory design and communication scheduling

𝒒𝒒 𝑡𝑡 : UAV trajectory; 𝜆𝜆𝑘𝑘 𝑡𝑡 : binary scheduling variable; 𝑇𝑇𝑡𝑡: completion time Expected communication throughput for user k:

UAV energy consumption:

ICC, 2019 120


Energy Minimization with Throughput RequirementEnergy-Efficient UAV Communication

UAV communicates with multiple users, each has a throughput requirement Minimize UAV energy, via joint trajectory design and communication scheduling

ICC, 2019 121

UAV energy

Throughput constraint

Max. speed constraint


Scheduling variables


Proposed SolutionsEnergy-Efficient UAV Communication

The problem is non-convex, infinite number of optimization variables Solution 1: Fly-hover-communicate protocol

Communication only while hovering Optimize hovering locations, visiting order, and flying speed Optimal speed: maximum-range speed Convex optimization to find hovering locations, and use Travelling Salesman

Problem with Neighborhood (TSPN) to find visiting order Solution 2: Joint design

Path discretization (as opposed to time discretization) Successive convex approximation (SCA)

ICC, 2019 122


Fly-Hover-Communicate ProtocolEnergy-Efficient UAV Communication

Hovering location for ground node 𝑘𝑘: �𝒒𝒒𝑘𝑘 Hovering/transmission time required: Hovering and communication energy:

Flying energy:

Total energy:

ICC, 2019 123


Ground node

Hovering location

Tradeoff between hovering-and-communication energy vs. flying energy

Joint Optimization with Path Discretization and SCAEnergy-Efficient UAV Communication

Fly-hover-communicate is strictly suboptimal since no communication while flying

The optimal solution for energy minimization needs to jointly solve the following problem

ICC, 2019 124


As the mission completion time is also a variable, use path discretization method

Then apply SCA to find a KKT solution


ICC, 2019 125

Target throughput: 100 Kbits Target throughput: 10 Mbits


Low throughput requirement: straight flight High throughput requirement: fly closer to ground nodes


ICC, 2019 126


Significant performance gain over benchmark schemes (hovering at geometric center or hovering above ground nodes)

UAV-Ground Energy Trade-offYong Zeng, The University of Sydney

127

Energy-Efficient UAV Communications

D. Yang, Q. Wu, Y. Zeng, and R. Zhang, “Energy trade-off in ground-to-UAV wireless communication via trajectory design,” IEEE Trans. Veh. Techn., July 2018.

ICC, 2019




Cellular-Connected UAV Field trials: from 2G to 4G 3GPP study Performance evaluation Advanced interference mitigation techniques QoS-aware trajectory design


ICC, 2019 128

Supporting Aerial Users: Field Trials from 2G to 4G

ICC, 2019 129

Yong Zeng, The University of SydneyCellular-Connected UAV: Field Trials

Approx. Time

Cellular Network Main Findings

2000’s 2G GSM GSM network can provide a usefulcomplementary channel for UAV

2011 3G UMTS 3G network could offer possible solution for non-safety-critical communications with moderate altitude

2016 4G LTE Feasibility of (semi-)autonomous UAV operation over LTE at low altitude

2016 4G LTE More detectable BSs as UAV moves higher, but also more severe interference


Cellular-Connected UAV: Field Trials

Cellular-Connected UAVs: Number of Detected Cells

Field measurements by Qualcom Strong signal strength for drone UES,

despite down-tilted BS antennas Number of detected cells generally

increases with altitude Expected due to free space propagation at

higher altitude

ICC, 2019 130

Qualcom, “LTE Unmanned Aircraft Systems”, Trial Report, v1.0.1, May 2017


Recent results by 3GPP

ICC, 2019 131

Yong Zeng, The University of SydneyCellular-Connected UAV: 3GPP Study

Interference detection Uplink interference mitigation

Uplink power control: e.g., altitude-dependent compensation factors 𝛼𝛼UE

Full-dimensional MIMO (3D beamforming) Directional antenna at UE

Downlink interference mitigation Intra-site joint transmission Enhanced initial access Coordinated data and control transmission

Mobility Refining handover parameters based on airborne status, path information…


Performance Evaluation (1/7)

ICC, 2019 132

Cellular-connected UAV in 3GPP urban macro (UMa) scenario 19 sites, each constituting 3 sectors (57 cells) Inter-site distance (ISD) 500 m, cell radius 166.7 m Two different array configurations: fixed pattern versus 3D beamforming

Y. Zeng, J. Lyu, and R. Zhang, “Cellular-connected UAV: potentials, challenges and promising technologies,” IEEE Wireless Communications, Feb. 2019.

Yong Zeng, The University of SydneyCellular-Connected UAVs: Performance Evaluation


ICC, 2019 133



ICC, 2019 134

BS antenna radiation pattern (fixed pattern configuration)

Antenna main lobe down-tilted towards ground Several side lobes with reduced lobe gain as moving away from main lobe



ICC, 2019 135

Scenario 1: UAV with dedicated channel Focus on one particular UAV with horizontal coordinate (250 m, 100 m) One channel is allocated exclusively (interference-free) Three UAV altitudes: 1.5 m, 90 m, and 200 m

Scenario 2: Shared channel by UAV and ground user Multi-user downlink communication with both UAV and ground users All users share the same channel Total users is 20, but with varying number of aerial users (UAVs) Horizontal locations uniformly distributed Ground users: altitude fix to 1.5 m Aerial users: altitude uniform between 1.5 m and 300 m



ICC, 2019 136

Fixed BS pattern: Low altitude: UAV likely associated with nearby cells (cells 1, 5, 9) High altitude: UAV likely to be associated with distant cells (via antenna

side lobe) 3D beamforming: likely associate with nearby cells for all altitudes



ICC, 2019 137

High altitude: less SNR variation (due to reduced scattering) Worst-case SNR increases at high altitude over ground: benefit of LoS link

compensates the weak gain of antenna side lobe 3D beamforming significantly improves the performance



ICC, 2019 138

Sum rate for different number of UAVs with shared channel Overall spectrum efficiency degrades as number of UAVs increases Expected since aerial users suffer from more severe interference due to LoS-

dominant channels


Inter-Cell Interference Coordination (ICIC) Cellular-Connected UAV: Interference Mitigation

139

W. Mei, Q. Wu, and R. Zhang, “Cellular-connected UAV: uplink association, power control and interference coordination,” submitted to IEEE Trans. Wireless Commun. Available on arXiv

Uplink transmission from a UAV to multiple BSs, under spectrum sharing with the existing ground users

Optimize the UAV’s BS associations and power allocations to maximize the weighted sum rate of UAV and existing ground users

ICIC designs Centralized ICIC: assume a central scheduler which is able to collect all required

information from all BSs in the ICIC region (theoretical benchmark) Decentralized ICIC: divide the cellular BSs into small-size clusters and UAV exchanges

information with cluster-head BSs only The decentralized ICIC achieves the performance close to the centralized one, but

with significantly reduced signalling overhead

ICC, 2019


Air-Ground Uplink NOMA

140

W. Mei and R. Zhang, “Uplink cooperative NOMA for cellular-connected UAV,” IEEE J. Sel. Topics Sig. Process, 2019.

Uplink transmission from a UAV to BSs, under spectrum sharing with the existing ground users

Conventional multiple access techniques OMA (TDMA, OFDMA) Non-cooperative NOMA (Local SIC only)

Proposed cooperative NOMA technique Some (idle) BSs with better channel

conditions are first selected to decode the UAV’s signals

The decoded BSs then forward the signals to their backhaul-connected BSs for interference cancellation

ICC, 2019

Cellular-Connected UAV: Interference Mitigation Yong Zeng, The University of Sydney

Multi-beam UAV Transmission

141

L. Liu, S. Zhang, and R. Zhang, “Multi-beam UAV communication in cellular uplink: cooperative interference cancellation and sum-rate maximization,” submitted to IEEE Trans. Wireless Commun. Available on arXiv

Coexistence of occupied and available GBSs

Each data stream is sent to multiple available GBSs via beamforming

One-hop backhaul links between occupied and available GBSs (e.g., X-haul) are utilized for cooperative interference cancellation

Considerable performance gain (e.g., DoF gain)

ICC, 2019

Cellular-Connected UAV: Interference Mitigation Yong Zeng, The University of Sydney

QoS-Aware Trajectory DesignCellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 142

Ubiquitous sky coverage is difficult at the initial phase for aerial support

Exploit the fully/partially controllable UAV mobility to avoid coverage holes

Example: UAV-enabled delivery, hard constraints only on the initial and final destination

BS coverage region and UAV path depends on QoS requirement


S. Zhang, Y. Zeng, and R. Zhang, “Cellular-enabled UAV communication: a connectivity-constrained trajectory optimization perspective,” IEEE Trans. Commun. Mar. 2019 (Invited Paper).

QoS-Aware Trajectory Design: Illustrative example Cellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 143


𝛾𝛾1 < 𝛾𝛾2

Coverage region

QoS-Aware Trajectory Design: System ModelCellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 144

UAV mission: Fly from pre-determined initial location 𝑈𝑈0 to final location 𝑈𝑈𝐹𝐹 QoS requirement: A minimum SNR �̅�𝜌 needs to be maintained with the cellular network Assumption: circular coverage area by ground BS


Problem Formulation

Cellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 145

Mission completion time

Minimize mission completion time, subject to QoS constraint

UAV initial location constraint

UAV final location constraint

QoS constraint

(P1) is non-convex and difficult to solve, since: QoS constraint is non-convex Trajectory 𝒖𝒖(𝑡𝑡) is a continuous function of 𝑡𝑡, involving infinite # of variables


Proposed SolutionCellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 146


Equivalent reformulation: minimize flying distance via optimizing GBS-UAV association sequence and handover location

Construct an undirected weighted graph based on the coverage regions of each GBS

Proposed SolutionCellular-Connected UAV: QoS-Aware Trajectory Design

ICC, 2019 147


Step 1: Construct an undirected graph

Step 2: Find GBS-UAV association sequence with shortest path algorithm, as if

UAV would fly on top of GBS

Step 3: Optimize handoff locations with obtained GBS-UAV association

sequence (convex problem)


ICC, 2019 148

Simulation Setup

𝑀𝑀 = 11 GBSs uniformly randomly distributed in a 10 km x 10 km square region

UAV altitude: 𝐻𝐻 = 90 m; GBS height: 𝐻𝐻𝐺𝐺 = 12.5 m

Maximum speed of UAV: 𝑉𝑉max = 50 m/s

Reference SNR at distance 𝑑𝑑0 = 1 m: 𝛾𝛾0 = 𝑃𝑃𝛽𝛽0𝜎𝜎2

= 80 dB

Benchmark trajectories:

Straight flight trajectory: UAV flies from 𝑈𝑈0 to 𝑈𝑈𝐹𝐹 in a straight line with

maximum speed 𝑉𝑉max Optimal trajectory: Optimal solution via exhaustive search over all feasible

GBS-UAV associations



ICC, 2019 149

Illustration of Trajectory Designs with �̅�𝜌 = �̅�𝜌max

Maximum achievable SNR target: �̅�𝜌max = 14.69 dB

Proposed trajectory similar to optimal trajectory

Straight flight trajectory cannot satisfy the given SNR target



ICC, 2019 150

Mission Completion Time versus SNR Target

Proposed trajectory design achieves near-optimal performanceMaximum SNR target for straight flight trajectory is 4.57 dB lower than the

proposed and optimal trajectory designs (10.12 dB vs. 14.69 dB)


Conclusions of Part II

ICC, 2019 151

Trajectory optimization: a new DoF for both UAV-assisted terrestrial communications and cellular-connected UAV TSP/PDP for (initial) path planning General framework with block coordinate descent and SCA Time or path discretization Shortest path algorithm, graph theory and convex optimization

Energy-efficient communication: critical due to limited onboard energy Fixed-wing and rotary-wing energy model Energy efficiency maximization given operation time Energy minimization given throughput requirement Throughput maximization given energy consumption

Cellular-connected UAV Feasible, but severe air-ground interference is the main bottleneck Practical BS cooperation via cooperate interference cancelation/NOMA 3D beamforming is promising


Outline

Introduction





ICC, 2019 152

UAV-Enabled MulticastingExtensions

153

Y. Zeng, X. Xu, and R. Zhang, "Trajectory design for completion time minimization in UAV-enabled multicasting“, IEEE Trans. Wireless Commun., April 2018.

UAV-enabled multicasting with random linear network coding: common information for multiple ground nodes

Applications: public safety communication, emergency response, video streaming, intelligent transportation systems

Minimize mission completion time, while satisfying target file recovery probability (each ground node needs to have a minimum “contact” duration with the UAV)

Optimal trajectory design is challenging (NP-hard): virtual BS placement for waypoint design travelling salesman for visiting order linear programming for optimal speed

ICC, 2019


UAV-Enabled Multicasting: Optimized Trajectory

154

Benchmark: strip-based trajectoryTotal traveling distance: 13.9 kmRequired time: 279 s

Optimized trajectoryTotal traveling distance: 8.1 kmRequired time: 173.5 s

Y. Zeng, X. Xu, and R. Zhang, "Trajectory design for completion time minimization in UAV-enabled multicasting“, IEEE Trans. Wireless Commun., April 2018.

ICC, 2019

Yong Zeng, The University of SydneyExtensions

UAV Communications with Caching

155

Critical issue in UAV communication: limited UAV endurance (e.g., ~30mins) Proposed solution: exploiting ground user caching Phase 1: UAV caches files to the selected ground users Phase 2: D2D file sharing among ground users Fundamental trade-off in file caching time and file retrieval time: joint trajectory

and caching design

X. Xu, Y. Zeng, Y. L. Guan, and R. Zhang, “Overcoming endurance issue: UAV-enabled communications with proactive caching,” IEEE JSAC, June 2018.

ICC, 2019

Extensions Yong Zeng, The University of Sydney

UAV Communications with Caching

156ICC, 2019


UAV-Enabled Wireless Power Transfer (WPT)Extensions

157

J. Xu, Y. Zeng, and R. Zhang, “UAV-enabled wireless power transfer: Trajectory design and energy optimization,” IEEE Trans. Wireless Commun., Aug. 2018.

A UAV-mounted mobile energy transmitter (ET) is dispatched to deliver wireless energy to a set of on-ground energy receivers (ERs) Facilitate large-scale implementation of

future WPT systems with improved energy transfer efficiency

Optimize UAV trajectory to maximize the minimum received energy among all ERs

Main techniques: Multi-location hovering by ignoring the

maximum UAV speed constraints Successive hover-and-fly via travelling

salesman and SCA-10 -8 -6 -4 -2 0 2 4 6 8 10

x(m)

-10

-8

-6

-4

-2

0

2

4

6

8

10

y(m

)

ER locations

Multi-location hovering

Successive hover-and-fly

SCP-based trajectory

ER 1 ER 2 ER 4

ER 6

ER 7

ER 9

ER 3

ER 5

ER 8

ER 10

ICC, 2019


Secure UAV-Ground Communications

158

Physical layer security via joint transmit power control and trajectory design

G. Zhang, Q. Wu, M. Cui, and R. Zhang, Securing UAV communications via joint trajectory and power control,” IEEE Transactions on Wireless Communications, Feb. 2019.

ICC, 2019


UAV-assisted Jamming

159

Opportunistic UAV jamming to assist in secure terrestrial communication

A. Li, Q. Wu, and R. Zhang, “UAV-enabled cooperative jamming for improving secrecy of ground wiretap channel,” IEEE Wireless Communications Letters, Feb. 2019.

ICC, 2019

Extensions Yong Zeng, The University of Sydney

Future Work

Cellular-Connected UAV: Promising Research Directions New architectures/protocols for cellular-connected UAVs Field measurement and prototype for LTE/5G-connected UAVs Channel measurement and modelling for UAV-BS communications Performance analysis for 3D aerial coverage Security and safety issues of cellular systems with UAVs Spectrum management and multiple access schemes Interference mitigation for cellular-connected UAVs Cellular systems with coexisting aerial and terrestrial users 3D beamforming MIMO/massive MIMO/mmWave technologies for cellular UAV systems Offline/online UAV trajectory design with QoS constraints Cellular-enabled UAV swarm Machine learning/AI-empowered UAV trajectory and communication design ……

160ICC, 2019


UAV-Assisted Communication: Promising Research Directions New protocols and use cases for aerial communication platforms Channel measurement and modelling for UAV-terminal and UAV-UAV links Multi-UAV cooperative communication/interference management UAV-aided information dissemination/data collection/offloading UAV communication with limited buffer size/energy/wireless backhaul Throughput-delay trade-off in UAV communications Physical-layer security in UAV-ground communications Energy-efficient joint communication and trajectory optimization UAV-ground energy/throughput trade-off UAV BS/relay 3D placement and dynamic movement UAV and ground BS coexistence and spectrum sharing MIMO/massive MIMO/mmWave for UAV-assisted communications UAV meets wireless power/energy harvesting/caching/edge computing Machine learning/AI for UAV communications and networking …..

161ICC, 2019

Yong Zeng, The University of SydneyFuture Work

Thank YouQ & A ?

ICC, 2019 162

Documents

Accessing from the Sky: UAV Communications for 5G and Beyond€¦ · Y. Zeng, Q. Wu, and R. Zhang, “Accessing from the Sky: a tutorial on UAV communications for 5G and beyond,’’