Genetic Algorithms Optimized Adaptive Wireless Network Deployment

Abstract

Advancements in UAVs have enabled them to act as flying access points that can be positioned to create an interconnected wireless network in complex environments. The primary aim of such networks is to provide bandwidth coverage to users on the ground in case of an emergency or natural disaster when existing network infrastructure is unavailable. However, optimal UAV placement for creating an ad hoc wireless network is an NP-hard and challenging problem because of the UAV’s communication range, unknown users’ distribution, and differing user bandwidth requirements. Many techniques have been presented in the literature for wireless mesh network deployment, but they lack either generalizability (with different users’ distributions) or real-time adaptability as per users’ requirements. This paper addresses the UAV placement and control problem, where a set of genetic-algorithm-optimized potential fields guide UAVs for creating long-lived ad hoc wireless networks that find all users in a given area of interest (AOI) and serve their bandwidth requirements. The performance of networks deployed using the proposed algorithm was compared with the current state of the art on several experimental simulation scenarios with different levels of communication among UAVs, and the results show that, on average, the proposed algorithm outperforms the state of the art by 5.62% to 121.73%.