Now a days, Sparse signal processing has evolved as an essential approach in multiple-input multiple-output (MIMO) systems, by facilitating an effective detection and estimation of sparse signals. In MIMO systems, the sparse signals have developed in several implementations, through incorporating a Channel Estimation (CE), signal detection, and interference mitigation. The built-in sparsity of these signals can be extracted to enhance system reliability, reduce computational complexity, and improve robustness to noise and interference, developing a sparse signal processing is a critical aspect of MIMO system design. Simultaneously, a Sparse Bayesian Learning (SBL) approach is proposed, which is trained to detect and estimate signal processing in MIMO systems. The proposed approach has been trained and evaluated on various scenarios, demonstrating its effectiveness in detecting and estimating sparse signals. The performance of the proposed SBL approach has been compared to existing methods like artificial intelligence detect (AIDETECT), and the proposed approach has been shown to achieve a better signal-to-noise ratio (SNR) of 16 dB, respectively.