Fractional Pilot Reuse and Allocation Strategy in Massive MIMO Systems

Abstract

Massive Multiple Input Multiple Output (M-MIMO) technology plays an important role in Fifth-Generation (5G) and beyond communication systems. It provides more benefits from enhanced Spectral Efficiency (SE) to improve energy efficiency and more consistency. These advantages are possible with accurate Channel State Information (CSI) present at the Base Station (BS). Verifying accurate CSI is difficult because of the required size of the coherence interval and the resulting limitations on pilot sequence length. So, Pilot Contamination (PC) is introduced when reusing the pilot sequences in nearby cells, which delays the SE enhancement. PC is presented as a bottleneck that limits the achievable throughput of multi-cell massive MIMO systems. In this work, a method for assigning pilot signals and improving pilot sequences is proposed using a Dynamic Attention-based Adaptive Autoformer (DA3) to reduce the effects of PC and increase the system's SE and throughput. The DA3-based Fractional Pilot Reuse (FPR) model classifies users as cell-center or cell-edge based on their Signal-to-Interference-plus-Noise Ratios (SINRs) values. Cell-edge users having lower SINR assign orthogonal pilots to reduce inter-cell interference, while cell-center users with higher SINR reuse the same pilots across different cells to enhance SE and throughput. The DA3 parameters, distance threshold, pilot reuse factor, and pilot allocation are optimized using the Secant Optimization Algorithm (SOA), which helps to increase the SE and throughput in the MIMO system. The developed SOA achieves near-optimal performance with lower computational complexity, thus making it highly suitable for interference management and eliminating the PC issues. The performance of the proposed model is tested with the classical pilot resource allocation methods to confirm its effectiveness.