中文摘要：

基于传统的QR分解算法,将最大似然算法应用到QR分解算法中,提出了一种改进的VBLAST解码算法,称为MLQR算法。该算法主要克服QR算法最先检测层性能差的缺点,将最大似然法应用到QR算法中,以最大限度地提高最先检测的两层信号的性能,然后回代到QR算法中检测余下层信号。仿真结果表明,提出的MLQR算法比QR算法和迫零算法在误码性能上要好,特别是在高信噪比下性能得到明显改善。

英文摘要：

An improved detection algorithm of VBLAST based on QR Factorization was proposed. In order to overcome the disadvantage of the poor performance of the first detected layer in QR algorithm, the maximum likelihood （ML） detection algorithm was applied to the QR factorization algorithm. The performance of the first detected two layers was improved. Then, the signals of the other layers were detected base on the first two layers’ signals using QR algorithm. The proposed algorithm was simulated and compared with the conventional detection algorithms. Simulation results show that the proposed MLQR algorithm is superior to the conventional QR algorithm and zero force （ZF） algorithm, especially with high SNR.