,
Research Scholar, Electrical Electronics and Communication Engineering, GITAM Deemed to be University , Bengaluru, Karnataka , India
Associate Professor, Electrical Electronics and Communication Engineering, GITAM Deemed to be University , Bengaluru, Karnataka , India
Large-Scale Heterogeneous Networks (HetNets) are needed to deliver high capacity and reliable connectivity, as wireless communication demands. Nevertheless, the high density and diversity of base stations in the settings pose a great challenge in terms of Energy Efficiency (EE), Load Balancing (LB), and Interference Management (IM). Conventional centralized architectures are usually not scalable and do not adjust to the dynamism of networks. This study suggests using EELBA-MARL (Energy Efficient Load Balancing and Interference Aware User Association in Large-Scale HetNets Using Distributed Multi-Agent Reinforcement Learning), which is a framework that is decentralized to be optimized in a scaled manner. With the help of the Distributed Multi-Agent Reinforcement Learning (MARL), the model allows base stations to be independent agents that learn the best user association and resource allocation policies, based on the local network conditions. As proven by the results of the experiment, EELBA-MARL gets better scores on all the most important measurements than the traditional benchmarks like IAW, PF, and IRA. The balance of the framework reached almost an ideal equilibrium with a Fairness and Load Balance of 0.989. It is also worth noting that it had an overall Total Power Consumption of 7.04 W, which considerably lowered the environmental impact with an overall Total Throughput of 4929.54 Mbps. The model was highly efficient in small cells with a high Pico Energy Efficiency of 3838.76 b/J, which is significantly better than the IAW (3610.29 b/J) and PF (3214.7 b/J) algorithms. The sensitivity and ablation experiments also confirmed the strength of the combined reward mechanism in the adjustment of interference reduction and energy conservation. These results verify that EELBA-MARL offers an efficient, scalable, and autonomous remedy in the upgrading of resource assignment and sustainable performance in next-generation 5G-and-beyond HetNets.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.
