基于SVD-K-means算法的软扩频信号伪码序列盲估计  

Blind estimation of pseudo-code sequence of soft spread spectrum signal based on SVD-K-means algorithm

作　　者：张慧芝 张天骐[1] 方蓉 罗庆予 ZHANG Huizhi;ZHANG Tianqi;FANG Rong;LUO Qingyu(School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065

出　　处：《系统工程与电子技术》2024年第1期326-333,共8页Systems Engineering and Electronics

基　　金：国家自然科学基金(61671095,61702065,61701067,61771085);信号与信号处理重庆市市级重点实验室建设项目(CSTC2009CA2003);重庆市自然基金(cstc2021jcyj-msxmX0836);重庆市教育委员会科研项目(KJ1600427,KJ1600429)资助课题。

摘　　要：针对通信中软扩频信号伪码序列盲估计困难的问题,提出一种奇异值分解(singular value decomposition,SVD)和K-means聚类相结合的方法。该方法先对接收信号按照一倍伪码周期进行不重叠分段构造数据矩阵。其次对数据矩阵和相似性矩阵分别进行SVD完成对伪码序列集合规模数的估计、数据降噪、粗分类以及初始聚类中心的选取。最后通过K-means算法优化分类结果,得到伪码序列的估计值。该算法在聚类之前事先确定聚类数目,大大减少了迭代次数。同时实验结果表明,该算法在信息码元分组小于5 bit,信噪比大于-10 dB时可以准确估计出软扩频信号的伪码序列,性能较同类算法有所提升。Aiming at the difficulty of blind estimation of pseudo-code sequence of soft spread spectrum signal in communication,a method of singular value decomposition(SVD)and K-means clustering was proposed.In this method,the data matrix of the received signal is constructed by non-overlapping segments according to the length of a periodic pseudo-code sequence.Secondly,the data matrix and the similarity matrix are respectively evaluated by SVD to complete the estimation of the size of the pseudo-code set,data noise reduction,rough classification and the selection of the initial clustering center.Finally,K-means algorithm is used to optimize the classification results,and obtain the estimated value of the pseudo code sequence.The algorithm determines the number of clusters before clustering,which greatly reduces the number of iterations.At the same time,the experimental results show that the algorithm can accurately estimate the pseudo-code sequence of the soft spread spectrum signal when the packet of information symbols is less than 5 bit and the signal to noise ratio(SNR)is greater than-10 dB,and the performance is improved compared with other algorithms.

关 键 词：软扩频信号 盲估计 奇异值分解 K-MEANS 

分 类 号：TN911.7[电子电信—通信与信息系统]