基于MMSE的近似最优Lattice Reduction辅助线性并行检测算法

A Near-optimal Lattice Reduction Aided Linear Parallel Detection Algorithm Based on MMSE

现有基于Lattice Reduction(LR)技术的多输入多输出(MIMO)系统检测算法,虽然可以有效地提高MIMO系统的误比特率(BER)性能,但其检测性能与最优的最大似然(ML)算法相比仍然存在差距。针对这一问题,提出了一种新的基于信道分组的线性Lattice Reduction辅助检测算法。该算法首先将信道分为两组,对通过条件最差子信道的信号采用最优的ML算法检测,然后将其从接收到的信号中消除,再采用Lattice Reduction技术对第2组信道进行优化,最终并行地对剩余信号进行检测。仿真结果表明:在16QAM(Quadrature Amplitude Modulation)和64QAM调制下,对于4×4的MIMO系统,该算法的误比特率性能达到了最优;对于6×6的MIMO系统,该算法相比最优的ML算法其检测性能相差不到0.5dB。

Existing multiple-input multiple-output（MIMO） detection algorithms based on Lattice Reduction（LR） can effectively improve the bit error rate（BER） performance.However,these detection algorithms have a large signal to noise ratio（SNR） gap when compared with the optimal maximum likelihood（ML） algorithm.In order to solve this problem,a new Lattice Reduction aided detection algorithm based on channel partition is proposed in this paper.In this algorithm,the signals through the worse conditional sub-channels are first detected with an ML algorithm.After cancelling the impact of these signals,the remaining are detected in parallel with the optimized sub-channels using Lattice Reduction.The simulation results show that,under 16QAM（Quadrature Amplitude Modulation） and 64QAM,the BER performance of the proposed algorithm can achieve the optimal result for a 4×4 MIMO system and have less than 0.5 dB SNR gap as compared with the ML algorithm for a 6×6 MIMO system.