MIMO系统中一种新颖的联合信道估计与检测算法

本文基于EM(Expectation-Maximum)算法,提出了一种简单而有效的联合信道估计与检测接收机结构。接收机中通过采用非穷尽列表形式的后验概率检测算法,避免了传统干扰抵消结构中各天线发送信号间的残余干扰对后验概率计算的影响。并进一步针对常用的非穷尽列表形式后验概率检测算法存在的列表冗余度大的问题,提出了自适应长度的列表球形译码算法(ASLSD,Adaptive Size List Sphere Decoding)。该算法通过更新检测半径和设置停止条件,使检测列表长度可随信噪比和迭代次数自适应变化。而且通过将列表操作与LSD(List Sphere Decoding)算法相结合,避免了符号序列在不同半径下的重复检测和排序操作。仿真表明,在复杂度方面,该算法需搜索的路径数远小于LSD算法。在算法性能方面,以3次迭代10~(-4)误码率为例,该算法与PIC算法相比可以获得近2dB的性能增益,因而具有更优的性能与复杂度的折衷。

A Novel Joint Channel Estimation and Detection Algorithm for MIMO Systems

In this paper,a practical receiver structure based on EM channel estimation algorithm is proposed.By exploiting the non-exhaustive list map detection,the interference from transmit signal of different transmitters is avoided.Then to further reduce the re- dundancy of the existing non-exhaustive list map detection algorithms,a modified LSD detection algorithm with adaptive list size is pro- posed.Through updating detection radius and setting stop criterion,the detection list has a variable length which is adaptive with the SNR and the iteration.Moreover,by combining LSD with a list,not only the repeated detection with different radius is avoided,but also the ordering procedure is left out.Simulation shows that the proposed algorithm only need much shorter list size compared with LSD and can improve the BER at 10~(-4)with three iterations about 2dB compared with PIC algorithm,thus has better tradeoff between performance and complexity.