We study auction-theoretic scheduling in cellular networks using the idea ofmean field equilibrium (MFE). Here, agents model their opponents through adistribution over their action spaces and play the best response. The system isat an MFE if this action is itself a sample drawn from the assumeddistribution. In our setting, the agents are smart phone apps that generateservice requests, experience waiting costs, and bid for service from basestations. We show that if we conduct a second-price auction at each basestation, there exists an MFE that would schedule the app with the longest queueat each time. The result suggests that auctions can attain the same desirableresults as queue-length-based scheduling. We present results on the asymptoticconvergence of a system with a finite number of agents to the mean field case,and conclude with simulation results illustrating the simplicity of computationof the MFE.
展开▼
