Matrix factoriza@on techniques for data clustering and radio map reconstruc@on with applica@ons to UAV placement Urbashi Mitra Ming Hsieh Department of Electrical Engineering University of Southern California, Los Angeles, USA with Jun8ng Chen (USC), Omid Esrafilian (Eurecom) & David Gesbert (EURECOM) Acknowledgements This research has been funded in part by one or more of the following grants: ONR N00014-15-1-2550, NSF CNS-1213128, NSF CCF-1718560, NSF CCF-1410009 , NSF CPS-1446901, and AFOSR FA9550-12-1-0215.
Explora8on-Exploita8on Autonomous Underwater Vehicle (AUV) Ummaned Aerial Vehicle (UAV)
AUVs and Communica8ons
AUVs as data mules ¨ Autonomous Underwater Vehicle (AUV) ¨ Sensors in a field ¨ AUV needs to collect data from sensors 1 Sensor 3 Dynamic source ¨ Underwater 0.5 communica@on Vehicle start Z (km) hinders data Trajectory 0 collec@on Sensor 2 − 0.5 Sensor 1 ¨ How to traverse 1 field? − 1 0 1 0.5 0 − 0.5 § Path planning problem − 1 − 1 X (km) Y (km)
UAV Relay Placement 5 Op@mize the posi@on of a drone for a dynamic network over complex terrain Device-to-device 400 feet connec@on, mmWave applica@on, etc. intelligent transporta@on 1 mile Search and Surveillance Key & Challenge: Know the shadowing & avoid it (More generally, learn the fine-grained environment & adapt to it)
How to Efficiently Learn Radio Map? 6 BS 1000 45 35 800 Building height [m] Latitude [m] 25 user 600 15 5 (source: Nokia) 400 400 600 800 1000 Longitude [m] Simula@on of a theore@cally Simulated SNR map in City model, top view simplified throughput map a more realis@c sedng Challenges: accurate modeling of buildings (materials) and incomplete data sets When a global radio map is available, op@mal UAV placement straigheorward
Is there an underlying model? 7 Probabilis@c model “Fingerprint” model 3D terrain-map based (Hourani et. al’14 TCOML, (Romero et. al’17 TSP) (Monserrat et. al. ’15 IJAP) Mozaffari et. al’16 ICC) Over simplified Least structured Not robust to High complexity missing data large bias
Recommend
More recommend
