Energy-Efficient Base Station Activating/Sleeping Strategies for Two-tier Heterogeneous Networks

In wireless communications, maximising the energy efficiency of a heterogeneous cellular network has attracted a lot of attention from researchers as it is required to guarantee the quality of service (QoS). We note that the probabilities of activating base stations (BSs) in conjunction with BS sleeping strategies, partial spectrum reuse (PSR), and bias factors in adjusting the power consumption of BSs have not been extensively studied in maximising the network energy efficiency. Besides, mobile users' mobility may lead to an unbalanced distribution of traffic load among BSs, which will affect the energy efficiency of the network. To address the above research gap, the main objective of this thesis is to propose solutions to maximising the energy efficiency of a two-tier network while considering BS activation probabilities, BS sleeping strategies, PSR, BS bias factors, and user mobility. The first contribution is to maximise the network energy efficiency through joint optimisation of the activation probability of small cell base station (small BS), the activation probability of macrocell base station (macro BS) and the PSR factor. The simulation results show that a higher activation probability of small BSs and a higher PSR factor leads to the network's higher energy efficiency, where the inter-tier interference is mitigated by applying PSR. The second contribution is to maximise the network energy efficiency by jointly optimising the spatial density of active BSs and the user-cell association indicators. Hence, a Switching Off Decision and User Association (SODUA) algorithm is proposed for a Control-Data Separation Architecture (CDSA) where it allows the macro BS to control the small BSs to be switched on or off. The simulation results show that for a given service area, there is an optimal number of active small BSs maximising the network energy efficiency, but the energy efficiency cannot be further improved by switching off more small BSs than a certain number. The third contribution considers four modes of small BSs: On, Standby, Sleep, and Off, where a different bias factor is associated with each mode, CDSA is considered for the macro BS to determine the mode of each small BS, and the macro BS is always in the On mode and is associated with the corresponding bias factor without having the four modes since it is always in active mode to control the network. A Genetic Algorithm based Power Mode Variant Selection (PMVS) algorithm is proposed to maximise the network energy efficiency by jointly optimising the bias factors of all BSs. The simulation results reveal that the proposed Genetic Algorithm based PMVS algorithm improves the two-tier network's energy efficiency as compared with the SODUA algorithm.