Mean Field Evolutionary Dynamics in Ultra Dense Mobile Edge Computing Systems

In mobile edge computing(MEC), the computing resources used to be centralized on the core cloud are extended to the mobile edge hosts (i.e. cloudlets) dispersedly deployed near the mobile users. In this new architecture, myriad of mobile terminals including vehicles, smart phones and different kinds of computers will form an ultra dense network. This motivates us to consider an ultra dense MEC system in which limited mobile edge hosts serve a relatively huge amount of mobile users. Considering the substantial amount and selfish manner of the users, we formulate the problem as a non-cooperative dynamic population game. The main contribution of this paper is as following. First, we consider the channel interference, average response time, load balance among servers and fairness (the variance of individual response time). Second, we propose an innovative mean field evolutionary approach which is robust to the channel interpolation as shown by the simulation results. Eventually, we consider a more challenging situation when the computing resources are rather limited.