Long-Term Contract Design for Traffic Off-Loading in Heterogeneous Cloud Radio Access Networks

Abstract

In order to improve the energy efficiency of the heterogeneous cloud radio access networks (H-CRANs), it is a promising approach to shutting off the small base stations (SBSs) with low traffic load and offload the traffic to the active SBSs nearby. Due to the selfish nature of the SBSs, the nearby SBSs may not cooperate if there is not appropriate incentive. On the other hand, since a user will be served by a SBS for a period of time, during which the SBS's offloading ability, i.e., its own traffic load and the offloading channel conditions, may vary dramatically. In such a case, how to provide proper long-term incentives for the potential SBSs to take over the traffic is an essential issue, especially when these SBSs belong to different service providers. Considering the SBS's offloading ability is the privat information that is unknown to the shutting off SBS, in this paper, the SBSs' cooperation stimulation problem is formulated as a long-term contract design problem, where the shutting off SBS acts as the principal and the offloading SBSs are the agents. A contract-theoretic framework based on the Markov decision process is formulated to study the long-term utilities of both parties. Through theoretical analysis, the feasible conditions for the optimal solution are described. Finally, the optimal contract is obtained through the proposed algorithm, and simulation results validate the effectiveness of our proposed long-term contract-based incentive mechanism for traffic offloading in H-CRANs.