基于半定松弛的长码DSSS信号扩频波形估计  

A Semidefinite Relaxation Approach to Spreading Waveform Estimation for Long-Code DSSS Signals

作　　者：张花国[1] 李鑫[1] 张建华 魏平[1] 

机构地区：[1]电子科技大学电子工程学院,四川成都611731 [2]中国船舶工业系统工程研究院,北京100094

出　　处：《电子学报》2016年第2期334-339,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61201282);中央高校基本科研业务费(No.ZYGX2013J016)

摘　　要：针对非合作通信中的长码DSSS信号,提出了一种基于半定松弛的扩频波形估计方法,并在确定信号模型下推导了扩频波形估计的CRB.首先推导了长码DSSS信号扩频波形的极大似然估计,由于该极大似然估计为非凸的组合优化问题,提出通过松弛约束条件将其转化为具有多项式计算复杂度的半定规划问题,实现对该极大似然估计问题的近似求解.仿真表明本文提出方法性能优于现有方法,并随着信噪比的提高逐渐逼近CRB.For long-code direct sequence spread spectrum（ DSSS） signals in non-cooperative communication systems,a semidefinite relaxation approach to spreading waveform estimation is proposed,and the Cramer-Rao lower bound（ CRB） for the spreading waveform estimation is also derived under the deterministic signal model. We first derive the maximum likelihood estimate（ MLE） of spreading waveform. Then,due to the MLE problem being a non-convex combinatorial optimization problem,we approximate it as a semidefinite programming problem which features polynomial worst-case complexity by relaxing the constraints. The simulation results demonstrate that the proposed estimator significantly outperforms the existing estimators and can achieve the CRB as the signal-to-noise ratio increases.

关 键 词：长码DSSS信号 极大似然估计 半定松弛 扩频波形估计 CRB 

分 类 号：TN911.6[电子电信—通信与信息系统]