Optimizing Stochastic Small Base Station Deployment with Particle Swarm Technique

Abstract

AbstractSmall base stations (SBSs) play a vital role in 5G communication to improve the throughput of cellular networks. However, care needs to be taken to ensure that improving the throughput of a cellular network via SBS deployment does not lead to unacceptable interferences that negatively impact the network’s overall efficiency. The unpredictable nature of SBS deployment also has implications for energy consumption. This research study proposes a weighted-sum modified particle swarm optimization (PSO) algorithm to find the density of SBSs that maximizes the throughput and energy gains of a cellular network. A stochastic geometry approach is taken to the optimization process, and some form of SBS sleep strategies are also explored at high and low traffic levels. The study showed that the strategic sleep mode favours lower densities of SBSs at lower transmission power levels than the random sleep mode at low traffic levels. The strategic sleep mode selects higher densities of SBSs at higher transmission power levels than the random sleep mode at high traffic levels. The strategic sleep mode provided a better optimal solution to the SE and EE maximization problem at both high and low transmit levels. The proposed PSO algorithm generated all Pareto optimal fronts regardless of the network traffic level. In contrast, the ParetoSearch algorithm could generate the Pareto optimal front at only low traffic levels. The result of this study provides cellular network engineers with a means of simultaneously adjusting network parameters to achieve the desired throughput and energy savings in SBS-enhanced cellular networks.KeywordsHeterogeneous cellular networksStochastic geometryParticle swarm optimizationSmall base stationMacrocellSmall cellThroughputUser equipment