Queue-aware energy-efficient scheduling in small-cell networks

Abstract

We investigate the problem of throughput-guaranteed energy saving for mobile stations (MSs) in small-cell networks (SCNs) with N base stations (BSs) and K MSs. We formulate the problem as a stochastic control optimization problem of minimizing the average total power expenditure in each time slot over infinite horizon when the data arrival rate vector is within the mean rate stable region. Then a dynamic scheduler which jointly considers power control and MS-BS matching in each time slot is proposed. The scheduler takes full advantage of both queue state information (QSI) and channel state information (CSI) for cross-layer opportunistic scheduling with coordinated multi-BSs. Based on the dynamic scheduler, we present an optimal algorithm with O(NK) complexity for power allocation and O[max (N,K) <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> ] complexity for user assignment via Hungarian algorithm. Simulation results show that our scheme could achieve better energy efficiency compared with Round Robin and MaxWeight algorithms as well as the RSS-based algorithm (GAPA). Meanwhile, our scheduler is able to ensure that the throughput of the system is optimal.