Extended jointly Gaussian approach for MIMO multipath channels

In this paper, an extension of the jointly Gaussian approach (JGA) to detect signals transmitted over a multiple-input multiple-output (MIMO) multipath channel is presented. This method, dubbed extended JGA, enhances the performance of the JGA by performing a marginalization over some interfering terms before applying the jointly Gaussian assumption to remove the residual interference. Hence, it can improve the performance of a communication system at the expense of a controllable increase in the computational complexity. The signal-to-noise-plus-interference ratio (SINR) at the detector's output is analyzed to indicate how much the extended JGA method can enhance the system's performance. Bit error rate (BER) curves are used to corroborate the gain in performance pointed by the SINR analysis for realistic MIMO scenarios.