NEXTWAVE ESTIMATOR: HYBRID INTELLIGENCE FOR 5G MASSIVE MIMO CHANNELS

Abstract

Massive Multiple-Input Multiple-Output (MIMO) systems are a cornerstone of 5G communications due to their high spectral efficiency and reliability. Accurate channel estimation remains a major challenge, especially under high mobility and interference conditions. Traditional model-driven techniques such as Least Squares (LS) and Minimum Mean Square Error (MMSE) provide fundamental performance but exhibit limitations in complex environments. Recent data-driven methods using statistical learning offer adaptability but lack physical interpretability. This paper proposes a hybrid model-driven and data-driven channel estimation framework that leverages the strengths of both approaches. The hybrid method uses physical channel models for initial estimation and machine learning algorithms to refine estimates under dynamic conditions. Simulation results demonstrate improved mean square error (MSE) and bit error rate (BER) performance compared to conventional methods. The proposed approach enhances estimation accuracy in 5G massive MIMO with reduced computational complexity. This framework supports robust channel estimation suitable for practical 5G deployments.