The Ad Hoc UAV Ground Network (AUGNet)

Transcription

1 The Ad Hoc UAV Ground Network (AUGNet) Timothy X Brown Interdisciplinary Telecommunications Electrical and Computer Engineering University of Colorado Presented at the AIAA 3 rd Unmanned Unlimited Technical Conference Chicago September 21, 24 CU-Boulder

2 Thanks to University of Colorado Brian Argrow Cory Dixon Jack Elson Sheetalkumar Doshi Daniel Henkel Jesse Himmelstein Sushant Jadhav Gerald Jones Marc Kessler Jake Nelson Phillip Nies Bill Pisano Roshan-George Thekkekunnel Institute for Telecommunication Sciences John Ewan Fidelity-Comtech Joe Carey L3-Corporation Ken Davey 9/21/24 T.X Brown ( 2

3 Unmanned Aerial Vehicles (UAVs) Small (1kg) Low-Cost ($1k) UAVs Swarming, flocking, cooperative missions Military, scientific, commercial applications 9/21/24 T.X Brown ( 3

4 Ad Hoc Networks A C B Network Layer Relaying Robust Connectivity 9/21/24 T.X Brown ( 4

5 Ad hoc UAV-Ground Networks Scenario 1: increase ground node connectivity. NOC Scenario 2: increase UAV mission range. 9/21/24 T.X Brown ( 5

6 We are building AUGNet COTS Radio Components Small Low-Cost UAV Full-Scale Test Bed. Fixed nodes Mobile ground nodes (vehicle/personnel carried) Aerial vehicle mounted nodes Test bed is another paper 9/21/24 T.X Brown ( 6

7 AUGNet Hardware Strategy Common hardware and software platform COTS communication hardware Open Source network software Custom UAVs Flexible and stable design 9/21/24 T.X Brown ( 7

8 Communication Hardware Environmental Enclosure 16cm UAV Mounting 21cm Common Hardware 82.11b PCMCIA card Soekris single board computer.1-1w Amplifier GPS (for monitoring) 256MB CF personality 9/21/24 T.X Brown ( 8

9 Communication Software F r o m K e rn e l F r o m D e v ic e (e th ) C 1 A 1 R o u te T a b le w ith S e n d B u ff e r B 2 R o u te R e q 2 3 D E 3 A R P T a b le B r o a d c a s t C la s s if ie r 1 C la s s ifie r O th e r P k t D is c a r d U n ic a s t C B A C K P k t A C K T a b le R e q u e s t T a b le H o s te th e r F ilte r 1 C la s s if ie r 3 O th e r E D D is c a rd C la s s if ie r 2 R o u te R e p ly D S R P k t O th e r C la s s if ie r 4 O th e r Linux operating system Click packet processing Dynamic source routing Self starting, automated No long-term state RAM file system ROM personality Robust Q u e u e 1 A D is c a r d T o K e rn e l T o D e v ic e (e th ) A 9/21/24 T.X Brown ( 9

10 Test Bed Table Mountain National Radio Quiet Zone Experimenting to show the effect of the UAV 9/21/24 T.X Brown ( 1

11 Delay P acket R o u n d T rip T im e (m sec) max average min hops Fixed nodes separated by 1 5 hops Delay ~1ms per hop Have tested VoIP application. Works well up to 3 hops 9/21/24 T.X Brown ( 11

12 Throughput Throughput Comparison 14 without UAV with UAV 12 Throughput/kbps "Nominal hops" Ad hoc throughput > 1kbps up to 5 hops UAV enables short cuts that preserve throughput 9/21/24 T.X Brown ( 12

13 Range T h r o u g h p u t (M b p s ) Grnd-Grnd UAV-Grnd % occurance Histogram Packet Loss in 2 sec Interval Packet Loss with UAV without UAV Distance (meters) UAV more than doubles the range UAV performance more erratic: more maneuvering Lost packets in % 9/21/24 T.X Brown ( 13

14 Latency/ms Node Mobility Latency - Mobile Nodes without UAV with UAV Within fixed node groups To/From Mobile Nodes Between fixed node groups Mobile Relay fixed nodes mobile src-dest fixed nodes with mobile relays Mobile nodes are not reliable relays UAV helps bridge unreliable links 9/21/24 T.X Brown ( 14

15 Summary: We are building COTS-based ad hoc networks on UAV, fixed, and mobile nodes. Range: 1.2km Grnd-Grnd, 3 km UAV-Grnd Low Delay, Good Throughput: Can do VoIP UAV maneuvering affects performance 9/21/24 T.X Brown ( 15

16 Next Steps: Coupling Plane Dynamics with Communication UAV Flocking Multi UAV operations Project website: augnet.colorado.edu 9/21/24 T.X Brown ( 16