Resilient Deployment of Drone Base Stations

Abstract

Drone base stations have emerged as a promising solution to the requirements of future cellular networks which may not be fully met by the existing terrestrial base stations. The main reasons for the superiority of DBSs are the higher probability of line of sight (LoS) and mobility in the sky which provides better adaptation to the demands of users. However, drones are complex electro-mechanical systems and more prone to the errors compared to that of radio communication systems. With the help of onboard sensors, failure tendency of a drone can be estimated and this information can be used to determine the positions of DBSs. In this work, we address the problem of DBSs' deployment where one of the DBSs is assumed to have a high probability of failure. Our proposed algorithm jointly determines the positions of DBSs before a failure occurs and paths to be followed in order to recover the network. Our simulations show that our proposed algorithm provides significant gain in the minimum user data rate performance of the network during recovery phase with tolerable loss in the initial performance compared to the benchmark algorithm.