Base station sleeping control and power matching analysis using extended G/M/1 queueing model

Abstract

This paper investigates the base station (BS) sleeping control and power matching using an extended G/M/1 queueing model with adjustable service rate, considering the general traffic arrival process. Besides the common Poisson arrival traffic model, some bursty traffic models, e.g. the Interrupted Poisson Process (IPP) and the On-Off Process, are also taken into account. The total power consumption and delay performance are analyzed. Especially, for the bursty traffic model, we explore the impact of traffic burstiness on the total power consumption and delay performance. Moreover, based on the delay tail distribution analysis, the delay-constrained total power minimization problem is solved, which provides the optimal energy-delay tradeoff performance.