Distributed Cooperation Under Uncertainty in Drone-Based Wireless Networks: A Bayesian Coalitional Game

Vandana Mittal,Setareh Maghsudi,Ekram Hossain

DOI: https://doi.org/10.1109/tmc.2021.3073772

IF: 6.075

2021-01-01

IEEE Transactions on Mobile Computing

Abstract:We study the resource sharing problem in a drone-based wireless network by considering a distributed control setting under uncertainty (e.g., due to lack of full information). The drones cooperate in serving the users while pooling their spectrum and energy resources in the absence of prior knowledge about different system characteristics such as the amount of available power at the other drones. Compared to the state-of-the-art research in drone-based wireless networks, which is mainly based on the assumption of accurate global information availability at every drone, our setting is realistic and practical. We cast the efficient resource pooling problem as a Bayesian cooperative game in which the agents (drones) engage in a coalition formation process, where the goal is to maximize the overall transmission rate of the network. The drones update their beliefs by using a novel technique that combines the maximum likelihood estimation with Kullback-Leibler divergence. We propose a decision-making strategy for repeated coalition formation that converges to a stable coalition structure. We analyze the performance of the proposed approach by both theoretical analysis and simulations. We provide the comparison of our scheme with the baseline and the socially optimal solution obtained from the exhaustive search. Simulation results demonstrate the superior performance of the proposed method in terms of the sum-rate of the network, the individual rate of the drones, and convergence properties.

computer science, information systems,telecommunications