Incorporating BS Sleeping Into Clustered Cell-Free Networking Improves Not Only Energy Efficiency but Also Sum Rate

Abstract

Clustered cell-free networking has appeared as a promising technique to overcome the cell-edge problem in cellular networks. This is achieved by dynamically decomposing the whole network into a number of non-overlapping weakly-interfered subnetworks and conducting joint processing among the base stations (BSs) in the same subnetwork. In each subnetwork, BSs far away from user equipments (UEs) contribute little to the total received signal strength. Switching those BSs into sleep mode would have little impact on the desired signal strength, while greatly reducing inter-subnetwork interference. Therefore, it can be expected that enabling BS sleeping in clustered cell-free networks could potentially improve both sum rate and energy efficiency. In this paper, we incorporate BS sleeping into clustered cell-free networking to maximize sum rate and energy efficiency. By decoupling the optimization problem into a UE assignment subproblem and a BS assignment subproblem, an alternating optimization based sum rate maximization (AO-SRM) algorithm and an alternating optimization based energy efficiency maximization (AO-EEM) algorithm are proposed. Simulation results demonstrate the effectiveness of the proposed AO-SRM and AO-EEM algorithms, and more importantly, corroborate that enabling BS sleeping improves not only energy efficiency but also sum rate of clustered cell-free networks. This is in sharp contrast to cellular and fully cooperative networks where a tradeoff between sum rate and energy efficiency is always observed.