Application of UAV for a Disaster-Resilient System

Unmanned aerial vehicles (UAVs) commonly known as drones have several emerging applications with the continuous development of their various features. It highly suits for disaster resilient applications considering ease of deployment, low implementation cost, and flexibility. UAVs and wireless communication can be discussed from two different angles: wireless communication for UAV-assisted application and UAV-assisted wireless communication. Both of them help to understand how it assists disaster-resilient applications. In this chapter, we discuss both the topics and a detailed analysis of UAV-assisted wireless communication for disaster-resilient applications. Initially, we discuss how the UAV and wireless communication assist each other for emerging future applications. Next, we present the various applications of UAV-assisted communication and the respective challenges to overcome. Finally, we have discussed a sample scenario to methodologically realize the problems related to the application of UAVs in disaster-resilient systems. The problem studies the placement and the user association of UAV aerial base station to temporarily enable the coverage in a region where the terrestrial communication infrastructure is destroyed due to a disaster. The objective of the problem is identified as maximizing the achievable rate while increasing fairness among the users. The problem is analyzed by dividing it into three subproblems. Those are user association and clustering subproblem, intra-cluster positioning subproblem, and the altitude selection subproblem. User association and clustering problems are addressed through k-means clustering and the Gale–Shapley algorithm. The intra-cluster placement subproblem is addressed via a modified pattern search algorithm. The altitude selection subproblem is solved by understanding the impact on the achievable rate with the elevation angle between the user and the UAV. The performance of our proposed approach is compared through numerical simulation with one of the benchmark approaches provided in the current literature. Simulation results witness that our proposed approach gives at least a 43% gain compared to the benchmark approach.