Maintaining wireless communication coverage among multiple mobile robots using fuzzy neural network

Abstract

This paper introduces a decentralized control scheme for deploying multiple mobile robots into an unknown realistic environment with the fluctuating signal propagation condition to establish and maintain desired wireless communication connections. A fuzzy neural network controller is designed for each robot, trained by the backpropagation algorithm, and applied to decide the motion for each robot, based on perceived wireless link quality. The simulation results show that the proposed scheme can establish and maintain effective communication coverage under the documented path loss exponents and uncertainties, and cause the average link quality to converge towards the desired range.