enhancing communications using uav based distributed … › images › 7 › 7a ›...

1
Enhancing Communications using UAV based Distributed Antenna Arrays Han Yan, Samer Hanna, and Danijela Cabric, UCLA Conclusions UAV based sparse array facilitates MIMO multiplexing with conventional infrastructure in LOS channel. Higher carrier frequency benefits from increased degree-of-freedom gain without losing SNR Challenges and investigation direction of utilizing swarm array are reviewed. Introduction Degree of freedom (DOF) of a wireless MIMO channel depends on Amount of multipaths (scattering and reflection) Antenna array aperture (wavelength normalized) Current MIMO system exclusively relies on multipaths Indoor environment & dense urban Line-of-sight (LOS) is “unfavorable” environment Large aperture is rarely investigated due to space & hardware concerns Many applications operate in LOS environment Data backhauling to infrastructures Air-to-air & air-to-ground communication Millimeter-wave (mmW) band where non-LOS paths are negligible due to severe penetration and reflection loss. Fig (right). Sum SE and SE per eigenmode with various number of conventional Tx infrastructure; Top and bottom figures have same color coding. Challenges & Investigation Directions References [1] H. Yan, and D. Cabric, “Compressive Initial Synchronization and Cell Discovery for Millimeter-Wave Networks,” to be submitted to IEEE Tran. Signal Process., Oct. 2018. [2] H. Yan, Samer Hanna, Kevin Balke, R. Gupta, and D. Cabric, “Software Defined Radio Implementation of Carrier and Timing Synchronization for Distributed Array (invited paper),” to be submitted to IEEE Aerospace Conference (AeroConf), Oct. 2018. [3] H. Yan, S. Chaudhari and D. Cabric, “Wideband Channel Tracking for mmWave MIMO System with Hybrid Beamforming Architecture,” (invited paper) in Proc. IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), Dec. 2017 . Distributed processing & data sharing Centralized array processing, e.g., estimate channel and forwards to a fusion center, v.s. distributed array processing. e.g., blind eigenspace estimation Fig. LOS channel tracking for distributed beamforming applications. Adaptive weight adjustment stabilize beamforming gain when unavoidable phase drift occur. Synchronization & self-localization of distributed mobile radio [1,2] Fully wireless intra-group freq. & timing sync. protocol Intra-group directional info. facilitates mmW/THz array for data sharing Direction info. helps self-localization, critical step in array geometry optimization Array geometry optimization Optimize UAV location for low rank channel in beamforming and high rank channel in MIMO multiplexing Error-robust process & channel tracking Unavoidable location vibration of clock phase drifting need to be considered. Our previous work focuses on sparse (LOS) channel tracking [3] Reconfigurable Distributed Array Enabling LOS MIMO Multiplexing Simulation settings Each color-face represents unique eigenmode Dashed curves show top and bottom 1% tier of SE (array geometry dependent) Improving degree of freedom in LOS MIMO Channel How UVA based array help? Large aperture; few reconfigurable elements 32-Element Critically Spaced Array 1024-Element Critically Spaced Array (32x Larger Aperture) 32-Element Uniformly Spaced Sparse Array (32x Larger Aperture) 32-Element Arbitrarily Spaced Sparse Array (32x Larger Aperture) Large aperture; large elements Single Wide Main-lobe Narrow (Grating) Lobes Single Narrow Grating-lobe (w/ controllable spatial emission) small aperture; few elements Single Narrow Main-lobe Large aperture; few elements Long Distance (>1000 meter or >10000) Array Processing (Decoding 1 Data Stream) Information Precoding stream 1 Short Distance (~meter or ~10) Array Processing (Decoding 3 Data Stream) Array aperture (~10) Conventional antenna array (/ spaced) Facilitates beamforming and power gain Application: beamforming enabled power gain Fixed access for data backhauling Antenna array for satellite communication Microwave Single Element v.s mmW Sub-Array System Overview Coherent combining at Rx extends coverage Coherent combining reduces transmit power Devices out of coverage zone Distributed beamforming UAV array facilitates MIMO multiplexing access and backhauling even in LOS scenarios Adjustable virtual sub- array architecture Rank 2 channel Rank 3 channel Distributed MIMO Degree of freedom improvement w/ conventional infrastructure UAV array facilitates range extension for sensor data collection UAV based swarm array system Each drone has single microwave radio or mmW sub-array Mobility of UAV facilitates array geometry reconfiguration Noise concerns: LOS propagation loss Friis equation – wavelength & antenna gain DOF concerns: LOS channel rank Angular resolution of sparse array Δ~1/(/) where is the aperture size Capability of resolving spaced antenna: ≈ tan Δ ~/ where is the link distance For the same channel rank improvement Short relieves required drone spacing Short supports higher comm. distance Other considerations High throughput intra-group data sharing Self-localization via intra-group angle info. Sub-array gain alleviates or even reverses propagation loss in mmW. Rank 2 Channel Rank 1 Channel Tx Antenna Placement (m) Tx Antenna Placement (m) Tx Antenna Placement (m) Tx Antenna Placement (m) Tx Antenna Placement (m) Fig. Example of the designed grating lobe facilitates MIMO multiplexing in LOS channel; 32 UAV based Rx and 4 Tx antenna at {-1.5,-0.5,0.5,1.5}. Main results Distributed UAV swarm array has great improvement in MIMO channel rank 28GHz suffers from severe pathloss, but UAV w/ 28GHz sub-array facilitates unprecedented spectral efficiency (SE) in LOS MIMO Up to 41% additional SE via position optimization Tx (conventional infrastructure) Tx array w/ variable antenna size Constantly spaced Tx array (c ) Tx power t =23dBm Rx (UAV swarm) 10MHz noise BW 4dB NF Ideal Rx MIMO processing 32 drones w/ RMS radius =50m channel =1000m distance LOS pathloss model (Friis eq.) Full CSI in TRx; Water-filling power allocation

Upload: others

Post on 30-May-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Enhancing Communications using UAV based Distributed … › images › 7 › 7a › CONIX_CORES1... · 2018-12-04 · Multi-Sensor Adaptive Processing (CAMSAP), Dec. 2017 . Distributed

Enhancing Communications using UAV based Distributed Antenna ArraysHan Yan, Samer Hanna, and Danijela Cabric, UCLA

Conclusions

• UAV based sparse array facilitates MIMO multiplexingwith conventional infrastructure in LOS channel.

• Higher carrier frequency benefits from increaseddegree-of-freedom gain without losing SNR

• Challenges and investigation direction of utilizingswarm array are reviewed.

Introduction

Degree of freedom (DOF) of a wireless MIMO channel depends on• Amount of multipaths (scattering and reflection)• Antenna array aperture (wavelength normalized)

Current MIMO system exclusively relies on multipaths• Indoor environment & dense urban• Line-of-sight (LOS) is “unfavorable” environment• Large aperture is rarely investigated due to space & hardware concerns

Many applications operate in LOS environment• Data backhauling to infrastructures• Air-to-air & air-to-ground communication• Millimeter-wave (mmW) band where non-LOS paths are negligible due to

severe penetration and reflection loss.

Fig (right). Sum SE and SE per eigenmode withvarious number of conventional Tx infrastructure;Top and bottom figures have same color coding.

Challenges & Investigation Directions

References[1] H. Yan, and D. Cabric, “Compressive Initial Synchronization and CellDiscovery for Millimeter-Wave Networks,” to be submitted to IEEE Tran. SignalProcess., Oct. 2018.[2] H. Yan, Samer Hanna, Kevin Balke, R. Gupta, and D. Cabric, “SoftwareDefined Radio Implementation of Carrier and Timing Synchronization forDistributed Array (invited paper),” to be submitted to IEEE AerospaceConference (AeroConf), Oct. 2018.[3] H. Yan, S. Chaudhari and D. Cabric, “Wideband Channel Tracking formmWave MIMO System with Hybrid Beamforming Architecture,” (invitedpaper) in Proc. IEEE International Workshop on Computational Advances inMulti-Sensor Adaptive Processing (CAMSAP), Dec. 2017 .

Distributed processing & data sharing• Centralized array processing, e.g., estimate

channel and forwards to a fusion center, v.s.distributed array processing. e.g., blindeigenspace estimation

Fig. LOS channel tracking for distributedbeamforming applications. Adaptive weightadjustment stabilize beamforming gain whenunavoidable phase drift occur.

Synchronization & self-localization ofdistributed mobile radio [1,2]• Fully wireless intra-group freq. & timing

sync. protocol• Intra-group directional info. facilitates

mmW/THz array for data sharing• Direction info. helps self-localization,

critical step in array geometry optimization

Array geometry optimization• Optimize UAV location for low rank channel

in beamforming and high rank channel inMIMO multiplexing

Error-robust process & channel tracking• Unavoidable location vibration of clock

phase drifting need to be considered.• Our previous work focuses on sparse (LOS)

channel tracking [3]

Reconfigurable Distributed Array Enabling LOS MIMO Multiplexing

Simulation settingsEach color-face representsunique eigenmode

Dashed curves show top andbottom 1% tier of SE (arraygeometry dependent)

Improving degree of freedom in LOS MIMO Channel

How UVA based array help?

Large aperture; few reconfigurable elements

32-Element CriticallySpaced Array

1024-Element Critically Spaced Array

(32x Larger Aperture)

32-Element UniformlySpaced Sparse Array

(32x Larger Aperture)

32-Element Arbitrarily Spaced Sparse Array (32x Larger Aperture)

Large aperture;large elements

Single Wide Main-lobe Narrow (Grating) Lobes

Single Narrow Grating-lobe(w/ controllable spatial emission)

small aperture; few elements

Single Narrow Main-lobe

Large aperture; few elements

Long Distance (>1000 meter or >10000𝜆)

Array Processing

(Decoding 1 Data Stream)

Information Precoding

stream 1

Short Distance (~meter or ~10𝜆)

Array Processing

(Decoding 3 Data Stream)

Array aperture (~10𝜆)

Conventional antenna array (𝝀/𝟐 spaced)• Facilitates beamforming and power gain• Application: beamforming enabled power gain• Fixed access for data backhauling• Antenna array for satellite communication

Microwave Single Element v.s mmW Sub-Array

System Overview

Coherent combining at Rx extends coverage

Coherent combining reduces transmit power

Devices out of coverage zone

Distributed beamforming

UAV array facilitates MIMO multiplexing access and backhauling even in LOS scenarios

Adjustable virtual sub-array architecture

Rank 2 channel Rank 3 channel

Distributed MIMO

Degree of freedom improvement w/

conventional infrastructure

UAV array facilitates range extension for sensor data collection

UAV based swarm array system• Each drone has single microwave

radio or mmW sub-array• Mobility of UAV facilitates array

geometry reconfiguration

Noise concerns: LOS propagation loss• Friis equation – wavelength 𝜆 & antenna gain

DOF concerns: LOS channel rank• Angular resolution of sparse array

Δ𝜙~1/(𝐴/𝜆) where 𝐴 is the aperture size• Capability of resolving 𝑑 spaced antenna:

𝑑 ≈ 𝑅tan Δ𝜙 ~𝜆𝑅/𝐴 where 𝑅 is the linkdistance

For the same channel rank improvement• Short 𝜆 relieves required drone spacing 𝐴• Short 𝜆 supports higher comm. distance 𝑅

Other considerations• High throughput intra-group data sharing• Self-localization via intra-group angle info.

Sub-array gain alleviates or evenreverses propagation loss in mmW.

Rank 2 ChannelRank 1 Channel

Tx Antenna Placement (m) Tx Antenna Placement (m)

Tx Antenna Placement (m)Tx Antenna Placement (m)

Tx Antenna Placement (m)

Fig. Example of the designed grating lobe facilitatesMIMO multiplexing in LOS channel; 32 UAV based Rx and4 Tx antenna at {-1.5,-0.5,0.5,1.5}.

Main results• Distributed UAV swarm array has great improvement in MIMO channel rank• 28GHz suffers from severe pathloss, but UAV w/ 28GHz sub-array facilitates unprecedented

spectral efficiency (SE) in LOS MIMO• Up to 41% additional SE via position optimization

Tx (conventional infrastructure)• Tx array w/ variable antenna size• Constantly spaced Tx array (∀𝑓c)• Tx power 𝑃t=23dBm ∀𝑓𝑐

Rx (UAV swarm)• 10MHz noise BW• 4dB NF• Ideal Rx MIMO processing• 32 drones w/ RMS radius 𝐴=50m

channel• 𝑅=1000m distance• LOS pathloss model (Friis eq.)• Full CSI in TRx; Water-filling

power allocation