Massive MIMO and LSTM-PD Algorithm for 5G Networks

Abstract

Research Paper on Massive MIMO and LSTM-PD Algorithm for 5G Networks Massive MIMO (Multiple Input and Multiple Output) is a key technology for enhancing the capacity and spectrum utilization of 5G (the 5th Generation) systems. This paper introduces the fundamental principles of Massive MIMO and proposes a novel LSTM-PD (Long Short-Term Memory with Periodic Decision) algorithm. The LSTM-PD algorithm, belonging to the category of periodic decision methods with predictive properties, is designed to optimize spectral efficiency and computational resource utilization. Additionally, a monitoring exit mechanism is incorporated to enhance the algorithm's performance. Experimental results on current network data demonstrate that when the physical resource block (PRB) utilization exceeds 30%, the spectral efficiency of the LSTM-PD algorithm significantly outperforms traditional algorithms. Moreover, when PRB utilization reaches 60%, the CPU utilization of the LSTM-PD algorithm is reduced by nearly 30% compared to traditional methods.