Nature Inspired Energy Optimisation of a Two-tier Network using Bias Factor

Abstract

The mobile users’ mobility causes some of the small BSs to be lightly loaded. The lightly loaded small BSs consume some amount of energy that can cause higher power consumption. To minimise the total power consumption, some of the small BSs need to be turned off but this would cause the existance of coverage holes. The motivation behind this work is to maximise the energy efficiency (EE) of a two-tier network that would lead to a green environment. Therefore, this paper investigates the EE of a two-tier network that comprises a control-data separation architecture (CDSA), where the macro cell base station (macro BS) provides full coverage while the small cell base stations (small BSs) provide high data rate services. The bias factors are used to adjust the power consumption of each BS’s mode in a power selection mode (e.g., On, Standby, Sleep, and Off). First, we model the spatial distributions of the small BSs and the mobile users following two independent Poisson point processes and derive the expressions for the downlink signal-to-interference ratio (SIR), a binary user-cell association indicator, power consumptions of the small BSs and the macro BS, as well as EE of the two-tier network. Based on the derived expressions, we formulate the EE maximisation problem and solve it by proposing a Genetic Algorithm (GA) based Power Mode Variant Selection (PMVS) algorithm. The GA is used due to its ability to obtain an optimal solution of multiple power selection mode parameters. It decides the appropriate mode for each small BS by using ranking that based on the nearest to the farthest location from the mobile users to the small BSs. The PMVS algorithm also calculates the EE where the bias factors modify the power consumption of the BSs subject to several constraints. Simulation results show that the proposed algorithm achieves a higher EE as compared to the previous work.