中文摘要：

为降低信道均衡算法的复杂度，提出了一种适于LDPC（Low—Density Parity—Check）编码MIMO（Multi—Input Multi，Output）信道的基于迭代式概率数据辅助（PDA：Probabilistie Data Association）的次优软输出判决反馈均衡算法。滑动窗PDA检测器输出外信息作为软输入软输出LDPC信道解码器的先验信息，实现信道均衡与信道解码的联合信息更新，克服传统判决反馈均衡器误差传播的缺陷。仿真表明，泼均衡算法性能要比V—BLAST迫零检测算法提高1—2dB，当Eb／No为3dB时，误比特率达到10^-5且算法复杂度仪为O（N^3），可获得较为满意的码间干扰消除效果。

英文摘要：

In order to reduce the complexity of equalization algorithm, a near-optimal algorithm based on softoutput DFE （Decision-Feedback Equalization） techniques with PDA （Probabilistic Data Association） algorithm applied for LDPC （ Low-Density Parity-Check） coded MIMO （ Multi-Input Multi-Output） systems is proposed. The extrinsic information is extracted from the PDA by sliding windowing equalization and is then used as the prior information in the soft-input soft-output LDPC channel decoding to realize information updating jointly. The simulations show that the equalization algorithm inproves the performance in SNR of about 1 - 2 dB with the BER can reach 10^-5 while the signal to noise ratio Eb/No = 3 dB compared to V-BLAST and the proposed algorithm can reduce greatly the ISI （Inter-Symbol Interference） at a low overall complexity of O（N^3 ）.