CAJ | 학술논문

在发射与接收端天线数目相当的大规模多输入多输出（ MIMO）系统中，线性检测算法的性能较差，而检测性能优异的检测算法复杂度偏高。针对该问题，提出增强型并行干扰消除算法（MMSE?ePIC），将搜索空间扩大到噪声增强方向矩阵的正交向量空间，同时利用近似对数似然方程值排序来减小复杂度。仿真结果表明，在误比特率为10－5时， MMSE?ePIC算法所需单位比特噪声功率谱密度比传统最小均方误差串行干扰消除（ MMSE?SIC）算法小2 dB，复杂度较MMSE?SIC降低了25％，在保持良好检测性能的前提下，复杂度明显降低，可以较好地实现大规模MIMO的上行检测。
재발사여접수단천선수목상당적대규모다수입다수출（ MIMO）계통중，선성검측산법적성능교차，이검측성능우이적검측산법복잡도편고。침대해문제，제출증강형병행간우소제산법（MMSE?ePIC），장수색공간확대도조성증강방향구진적정교향량공간，동시이용근사대수사연방정치배서래감소복잡도。방진결과표명，재오비특솔위10－5시， MMSE?ePIC산법소수단위비특조성공솔보밀도비전통최소균방오차천행간우소제（ MMSE?SIC）산법소2 dB，복잡도교MMSE?SIC강저료25％，재보지량호검측성능적전제하，복잡도명현강저，가이교호지실현대규모MIMO적상행검측。
In MIMO systems,the linear detection algorithm is inferior when the number of receiving antennas is close to that of transmitting antennas,and the algorithms with good performance are usually of high complexity.To solve the problem,an MMSE?ePIC is proposed,where the searching space is expanded to normal eigenvector space of noise enhanced direction matrix,and the complexity is reduced based on the sorting of approximated log likelihood function values.The computer simulations show that the noise power spectral density of the proposed algorithm is 2 dB less than that of conventional MMSE?SIC,with decreased computational complexity of 25%. The modified MMSE detection algorithm has good detection performance and significantly decreased complexity,and approves to be a good option to implement uplink multi?user detection in large?scale MIMO communications.