Game Theoretical Bandwidth Allocation in UAV-UGV Collaborative Disaster Relief Networks.

Bincheng Ying,Zhou Su,Qichao Xu,Xiandong Ma
DOI: https://doi.org/10.1109/hpcc-dss-smartcity-dependsys53884.2021.00223
2021-01-01
Abstract:After a natural disaster, cellular communications may be hard to meet the needs of disaster relief due to the destruction of infrastructures, such as base stations and power stations. In order to transmit rescue data in time, the disaster relief network based on unmanned ground vehicles (UGVs) is advocated to transmit rescue-critical data to the relief center. However, due to the driving-difficult roads in the disaster area and small communication coverage of UGVs, UGVs-assisted dis- aster relief network faces the challenge to quickly and accurately collect wide-range disaster information. To tackle this problem, in this paper, we first take unmanned aerial vehicles (UAV s) to develop a UAV-UGV collaborative disaster relief network architecture, where the UAV s with high mobility, wide view and agility can efficiently help UGVs to perceive the disaster information. Then, to realize the efficient data delivery between UAVs and UGVs with limited bandwidth resources, we devise a game-based bandwidth allocation framework. In specific, the multi-leaders multi-followers Stackelberg game is used to model the interactions between UGVs and UAVs, where each UGV acts as a leader to determines its bandwidth capacity and each UAV acts as a follower to determines its access selection based on the bandwidth capacity. Especially, the non-cooperative game model is introduced to analyze the competitions among UGVs, and the evolutionary game is utilized to study the strategy adjustments of UAV s. Through the backward induction method, the optimal access selection of each UAV and the optimal bandwidth capacity of each UGV are obtained, which form the Stackelberg equilib-rium. At last, extensive simulation results demonstrate that the proposed scheme outperforms the conventional counterparts on the utilities of both UAVsand U G V s.
What problem does this paper attempt to address?