Biased Channel Gain Approach for Energy-Efficient UE-to-BS Association in Massive MIMO HetNet

Typically, energy-efficient operation of wireless networks is achieved by allocating resources, such as physical resource blocks (PRBs), transmit power, and number of base station (BS) antennas, in an energy-efficient manner. However, the allocation of these resources is tightly coupled with user (UE)-to-BS association, which impacts the performance and efficiency of the network. Therefore, in this work we focus on energy-efficient UE-to-BS association for multi-cell multi-user massive multiple input multiple output (mMIMO) heterogeneous networks (HetNets) and propose two bias-based UE-to-BS association algorithms. Firstly, a range expansion bias (REB) model is employed to increase the effective coverage area of each pico BS (PBS), improving overall network performance and energy efficiency (EE). Secondly, the sum of channel gains as a metric for UE-to-BS association is leveraged, seeking to maximize the total channel gain across UEs and BSs by prioritizing stronger connections. This approach aims to optimize the distribution of UEs among available BSs, thereby enhancing the quality of service provided to UEs. Our results, based on numerical system-level simulations describing a realistic metropolis (Berlin) scenario, demonstrate that the individual REBs (iREBs) per PBS lead to significant improvements in EE, while the channel-gainbased approach effectively boosts overall network throughput and EE, outperforming the greedy maximum channel gain, a baseline REB based, and the iREB based UE-to-BS association methods.