Near optimal MLD receiver for multiple input multiple output systems

In this paper, we propose a new group-based detector (GBD) to guarantee low complexity and high performance in multiple-input multiple-output systems. The proposed GBD is composed of whitener, interference cancellation, and a linear detector. The proposed GBD firstly detects symbols with the highest received signal-to-noise ratio to eliminate error propagation. For this, whither is applied to detect more reliable symbols and enhance performance. In addition, a linear detector is considered to detect the remaining symbols and reduce the computational complexity. Consequently, the proposed GBD provides the optimum maximum likelihood detector (MLD) while requiring significantly reduced complexity resources compared with the optimal MLD.