Accurate Direction of Arrival Estimation for 5G MIMO Sparse Channel Using Advanced Multikernel Based NNSBL

Accurate estimation of the Direction of Arrival (DOA) in MIMO communication systems is essential for efficient signal analysis and noise cancellation, especially in challenging conditions. Traditional methods often struggle with computational inefficiencies and degraded performance in complex environments. This paper addresses these challenges by introducing an innovative approach that integrates antenna reconfiguration with an advanced Kernel-based Non-Negative Sparse Bayesian Learning (NNSBL) algorithm, applied to a predefined grid. The method leverages sparse representations and Bayesian principles, enhanced with multi-kernel techniques, to improve adaptability to the complexities of MIMO channels. Simulation results demonstrate that the proposed method achieves superior performance, particularly at high Signal-to-Noise Ratios (SNRs). RMSE values are reduced from 1.75 for conventional NNSBL to 1.43 for the advanced multi-kernel NNSBL at optimal configurations. Additionally, the spectral analysis shows sharper and more accurate peaks at higher SNR values (20 dB to 30 dB), with noticeable improvements in DOA pinpointing compared to traditional techniques. These findings highlight the method's capability to adapt effectively to complex MIMO channel dynamics, offering reliable and computationally efficient solutions tailored to modern 5G wireless communication systems.