摘要
本文提出适用于MIMO系统中,基于对数似然比排序进行干扰消除的检测算法。本算法首先计算成对符号的对数似然比,根据似然比逐行计算对应的判决率,并对各行的判决率从小到大排序进行逐行消除。通过为成对符号建立对数似然比,可以根据最大后验准则判决提供可靠的排序信息。仿真结果显示:与典型的消除算法相比,基于对数似然比消除的算法,不但可以降低由于错误消除引起的差错扩散,而且相比于经典的MMSE、ZF干扰消除算法,其误码性能得到较大的改善。
In this paper,we propose a detection algorithm based on the log-likelihood ratio(LLR),which can cancel the noise and interference under MIMO system.The LLR detection algorithm can provide the reliability information on the maximum a posteriori(MAP) probability decision by computing the pairwise LLR for a couple of symbols.During the cancellation stage,LLR ordering method computes the probability of symbol error for every row vector,and executes the cancellation process from the row that has the smallest symbol error probability.Simulation shows that the proposed LLR-based ordering significantly outperforms conventional ordering algorithm.
出处
《电路与系统学报》
北大核心
2013年第2期62-66,共5页
Journal of Circuits and Systems
基金
国家自然科学基金项目(61072008)
北京市优秀人才培养资助计划(2011D005002000005)
北方工业大学青年重点科研基金