Joint multi-user resource scheduling and computation offloading in small cell networks

In this paper, we address computation offloading problem from mobile users to their serving small cell base stations. These base stations can be endowed with some computational capabilities providing thus users proximity access to the cloud services. We aim to jointly optimize the radio resource scheduling and computation offloading in order to minimize the average energy consumed by all the users terminals to process their mobile applications under average delay constraints tolerated by these applications. We investigate for this problem offline and online dynamic programming approaches and we devise deterministic solutions to find the optimal scheduling-offloading policy. The proposed solutions select only one user for scheduling, hence offloading, and decides for the other users either local processing or staying idle according to their application rates. We show that the offline strategy is optimal in terms of energy saving compared to the online strategy. It can benefit from prior knowledge on the channel statistics and the application properties to satisfy the users requirements.