Graduate School and Research Center in Digital Sciences

Learning to rest: A Q-learning approach to flying base station trajectory design with landing spots

Bayerlein, Harald; Gangula, Rajeev; Gesbert, David

ASILOMAR 2018, 52nd Asilomar Conference on Signals, Systems and Computers, 28-31 October 2018, Pacific Grove, USA

Abstract--We consider the problem of trajectory optimization for an autonomous UAV-mounted base station that provides communication services to ground users with the aid of landing spots (LSs). Recently, the concept of LSs was introduced to alleviate the problem of short mission durations arising from the limited on-board battery budget of the UAV, which severely limits network performance. In this work, using Q-learning, a model-free reinforcement learning (RL) technique, we train a neural network (NN) to make movement decisions for the UAV that maximize the data collected from the ground users while minimizing power consumption by exploiting the landing spots. We show that the system intelligently integrates landing spots into the trajectory to extend flying time and is able to learn the topology of the network over several flying epochs without any explicit information about the environment.

Document Doi Bibtex

Title:Learning to rest: A Q-learning approach to flying base station trajectory design with landing spots
Type:Conference
Language:English
City:Pacific Grove
Country:UNITED STATES
Date:
Department:Communication systems
Eurecom ref:5761
Copyright: © 2018 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Bibtex: @inproceedings{EURECOM+5761, doi = {http://dx.doi.org/10.1109/ACSSC.2018.8645103}, year = {2018}, title = {{L}earning to rest: {A} {Q}-learning approach to flying base station trajectory design with landing spots}, author = {{B}ayerlein, {H}arald and {G}angula, {R}ajeev and {G}esbert, {D}avid}, booktitle = {{ASILOMAR} 2018, 52nd {A}silomar {C}onference on {S}ignals, {S}ystems and {C}omputers, 28-31 {O}ctober 2018, {P}acific {G}rove, {USA}}, address = {{P}acific {G}rove, {UNITED} {STATES}}, month = {10}, url = {http://www.eurecom.fr/publication/5761} }
See also: