摘要
为实现频率选择性衰落信道下未知信道信息时的盲信源提取,建立一个等价的卷积混叠信号模型,利用观测数据的“过去”、“现在”样本空间,在时域中推导出一种基于四阶累计量的盲信号提取算法,避免了频域的“排序”问题.通过接收阵元的约束性布位,简化了运算,计算机仿真实验验证了算法的有效性,并表明其具备一定的噪声抑制能力.
An equivalent convolution mixture signal model is built to realize the blind source extraction without channel state information in frequent selective channel. It uses the ＂past＂ and ＂present＂ sample space of observed data to derive the blind signal extraction method based on fourth order statistic which avoids the permutation problem. Computation complexity is reduced by the constraint distribution of receiving elements. Simulation results prove the effectiveness and noise interference ability of the algorithm.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2007年第5期761-764,共4页
Journal of Harbin Institute of Technology
基金
国家自然科学基金资助项目(60572039)
关键词
多径信道
盲信源提取
卷积混叠
四阶累计量
奇异值分解
muhipath channel
blind source extraction
convolutive mixture
fourth order statistic
singular value decomposition
