The energy consumption of mobile networks, particularly the 5G Radio Access Network (RAN), is becoming a growing concern due to its environmental and economic implications. As the demand for higher data rates and low-latency services intensifies, 5G networks, integrating macro cells and small cells, are emerging as critical infrastructures. Although small cells improve coverage and capacity, their increased deployment could lead to a significant rise in the overall power consumption of 5G systems.
Current small cell selection strategies by User Equipment (UE), although effective in some cases, do not fully account for the dynamic nature of traffic conditions and the specific data requirements of users. Moreover, current techniques such as the maximum Signal-to-Interference-plus-Noise Ratio (max-SINR) and Cell Range Expansion (CRE) purely consider the signal strength of the link between the user and the base station to allocate users to the base station. However, this leads to inefficient utilization of base station resources and uneven distribution of load, causing congestion at some base stations while leaving others underutilized.
In order to address these gaps, this thesis proposes a Traffic Distribution Orchestrator (TDO) to manage the distribution of users between cells dynamically, and optimize energy efficiency without compromising network performance. The proposed cell selection model developed in this thesis also accounts for user mobility and dynamic traffic conditions. The model estimates instantaneous power consumption and informs a real-time algorithm user equipment-base station (UE-BS) association algorithm to dynamically allocate users to the cell which will enhance the energy efficiency of the network while ensuring the required Quality of Service (QoS) requirements. Complementing this, an adaptive sleep mode mechanism puts underutilized small cells in a low power mode and reactivates them when demand rises, using hysteresis to prevent state flapping and reduce idle power.
Through MATLAB simulations, the effectiveness of the model and algorithm is validated, with results indicating a significant reduction in network power consumption in heterogeneous 5G deployments. The proposed UE-BS association algorithm is compared with the max-SINR, CRE and a representative association method from the previous studies, whereas the proposed adaptive sleep mode mechanism is compared with fixed threshold sleep mode mechanism under both bursty and steady traffic. The proposed UE-BS association algorithm combined with the adaptive sleep mode mechanism reduces total network power consumption relative to baseline strategies. This research contributes to the advancement of sustainable 5G network architectures and offers insights into energy efficiency optimization in real-world scenarios.