Computation Offloading Management for MEC Assisted UAV Networks: A Multi-leader Multi-follower Game Approach

Abstract

Unmanned aerial vehicles (UAVs) have received extensive applications due to their high mobility and maneuverability. However, the limited storage, computing and battery resources of UAVs, significantly impedes their applications in computing intensive scenarios. Offloading computation tasks to edge service providers (ESPs) has shown to be a promising solution. In this paper, we tackle the computing offloading problem of UAVs to achieve the best possible tradeoff between ESPs' revenues and UAVs' resources demand. We focus on a scenario where each UAV can offload the computation tasks to multiple ESPs. To obtain ESPs' optimal pricing strategies and UAVs' optimal resources demand, we model the interaction between UAVs and edge servers as a Multi-leader Multi-follower Stackelberg game. Then, we solve the equilibrium of the game iteratively with Lagrange multiplier method. The simulation results show that the more resources the ESPs own, the higher utility of UAVs can obtain, and when the resources of one ESP are large enough, other ESPs will withdraw from the market.