Energy efficient multiuser scheduling: Statistical guarantees on bursty packet loss

In this paper, we consider energy efficient multiuser scheduling. Packet loss tolerance of the applications is exploited to minimize average system energy. There is a constraint on average packet drop rate and maximum number of packets dropped successively (bursty loss). A finite buffer size is assumed. We propose a scheme which schedules the users opportunistically according to the channel conditions, packet loss constraints and buffer size parameters. We assume imperfect channel state information at the transmitter side and analyze the scheme in large user limit using stochastic optimization techniques. First, we optimize system energy for a fixed buffer size which results in a corresponding statistical guarantee on successive packet drop. Then, we determine the minimum buffer size to achieve a target (improved) energy efficiency for the same (or better) statistical guarantee. We show that buffer size can be traded effectively to achieve system energy efficiency for target statistical guarantees on packet loss parameters.