Multi-leader Multi-follower Stackelberg Game Based Dynamic Resource Allocation for Mobile Cloud Computing Environment

摘要

Emerging mobile cloud computing (MCC) technology offers great potential for the mobile terminals to support running highly sophisticated applications. However, supporting such applications needs efficient resource management from cloud servers and mobile terminals. Moreover, the resource management technologies based on quality of experiences (QoE) can meet the demand of end users in a better way than those based on quality of services. In this paper, we address the resource management problem of MCC network with an acceptable QoE. Based on multi-leader multi-follower two-stage Stackelberg game model, the proposed scheme maximizes the utility function of MCC networks. Considering the scenario in which the cloud servers and mobile terminals are selfish to maximize their own interests, the network performance is greatly affected by their greediness. To achieve better network performance, control decisions are coupled with one another. Utility function considers not only the spectral efficiency and user satisfaction in the mobile terminal but also the pricing information in the cloud. Our proposed scheme can obtain a well-balanced performance between mobile terminals and cloud servers. In addition, the existence of Nash equilibrium in the proposed scheme is investigated. Theoretically, the maximum and minimum selling prices of bandwidth are deduced. Simulation results show that the effectiveness of the proposed algorithm and our proposed scheme outperforms equal allocation scheme in terms of the user satisfaction and network revenue.