Privacy-aware pilot scheduling and channel estimation for 6G ultra-massive mimo systems

The rapid evolution of 6G wireless communication demands highly secure and accurate Channel Estimation (CE) in ultra-massive Multi-Input Multi-Output (MIMO) systems. However, selecting an optimal Pilot Contamination Mitigation (PCM) strategy in conventional MIMO systems remains challenging, degrading CE performance. Therefore, this paper presents a novel privacy-aware PCM-based CE framework for 6G- enabled ultra-massive MIMO systems using Fuzzy Adaptive Inference Normalization System (FAINS) and Deep-Linearized Multi-Noise Training-Linear Swish long Short-Term Memory (Deep-LMNT-LS 2 TM). The proposed model begins with user scheduling. Then, the input message bits undergo data encoding, followed by Orthogonal Frequency Division Multiple Access (OFDMA) modulation. Next, orthogonal pilot sequences are designed, and secret keys are generated using Diffie–Hellman key exchange. Further, watermark embedding is done. After that, the Pilot Configuration Modes (PCMo) are selected using FAINS, followed by Inverse Fractional Fourier Transform (IFrFT) conversion. Additionally, Intelligent Reflective Surfaces (IRS) amplify signals, enhancing reflection. Here, the dataset undergoes pre-processing and feature extraction. Then, the CE is done using Deep-LMNT-LS 2 TM. Lastly, the signal undergoes inverse transformation, pilot extraction, demodulation, and secure decoding to reliably reconstruct the original data. Experimental results demonstrate that the proposed framework significantly achieves a Mean Squared Error (MSE) of 0.16, outperforming classical methods while offering enhanced scalability, security, and efficiency.