This study presents the optimization of energy-efficient resource allocation in massive Multiple-Input Multiple-Output (MIMO) systems that employ hybrid energy harvesting techniques, integrating renewable energy sources such as solar and wind with traditional power grids. The primary objective is to maximize energy efficiency and system throughput while addressing the operational challenges inherent in managing large-scale MIMO systems. The research introduces both offline and online optimization algorithms, utilizing advanced methods such as Markov chains and time-detection strategies to accurately predict and optimize energy dynamics and channel states. The offline algorithms, developed using fractional programming techniques, assume ideal conditions with complete channel state information, achieving an energy efficiency of 92% and a throughput of 1.8 Gbps, serving as a performance upper bound. Conversely, the online algorithms, designed to operate in real-time scenarios by leveraging statistical data and future state predictions, demonstrate an energy efficiency of 88% and a throughput of 1.75 Gbps. These online methods balance accuracy and computational complexity, providing a robust solution for sustainable wireless communication systems. The results confirm that the integration of hybrid energy sources, combined with advanced optimization algorithms, significantly enhances the performance and sustainability of massive MIMO networks, making a substantial contribution to the advancement of green communication technologies.
