Fast Widely Symbol Detection For Mimo Systems

Multiple input multiple output (MIMO) scheme has the potential of providing significant performance gains to meet the rapidly growing demand of high quality of communication service. To further enhance the performance of conventional V-BLAST based MIMO system and reduce its computational complexity, a fast widely symbol detection scheme is proposed in this paper. The proposed method incorporates the widely signaling scheme to widen the signal space so that performance enhancement can be achieved. However, the widely signaling scheme requires the received signal and its corresponding complex conjugate being stacked together, leading to the increase of computational complexity for symbol detection. Hence, to concurrently receive performance gains and also save computations, the proposed scheme is conducted recursively as follows. In each recursion, the order of the column vectors in the detection matrix is permuted according to the metric of SNR maximization, followed by the estimation of the corresponding transmit symbols using the permuted detection matrix, and completed by removing the signal counterpart, relevant to the estimated transmit symbols, from the received signal for interference cancelation. Thereafter, the detection matrix is updated iteratively for saving computations to avoid directly computing the matrix inversions. The recursive procedure is repeated and stops until all symbols are transmitted. From the results of computer simulation and complexity analysis, the proposed scheme outperforms the conventional layer detection method, such as V-BLAST, with even lower computational complexity.