摘要
针对BOC长码扩频信号的盲估计问题,提出了一种酉矩阵变换方法对组合码(扩频序列和副载波序列的组合)进行估计,该方法可以消除矩阵与信号子空间的相位模糊关系。首先将信息序列与组合码序列的长码直扩模型建模于虚拟多用户模型;然后计算信号的自相关矩阵,并通过搜索该矩阵Frobenius范数最大值的位置来确定接收信号的失步点;接着将接收信号连续采样,按组合码周期将信号分段,对各个分段进行自相关运算,对得到的自相关矩阵进行累加平均后,进行矩阵分解;最后利用酉矩阵变换方法确定信号子空间与特征向量的关系,得到出各个信息位对应的组合码部分,对各个部分的组合码进行结合可以得出完整的组合码序列。仿真分析表明,该方法能够在较低的信噪比下达到较精确的估计性能。
For the blind estimation of BOC long-code direct sequence spread spectrum signals,a combination code （combination of long-code spread spectrum sequence and subcarrier sequence） estimation method according to the unitary matrix which can eliminate the phase fuzzy relation between matrix and the signal subspace was presented.And long-code direct sequence spread spectrum signals based on long-code spread spectrum sequence and subcarrier sequence was equal to the virtual multi user mode.Firstly the Frobenius norm of autocorrelation matrix was exploited to estimate the synchronous offset.Then according to the combination code period,the continuous sampling signal was sectioned,calculating singular value decomposition of the average of each autocorrelation matrix.Finally the relationship between the signal subspace and feature vector was determined using the unitary matrix transformation to eliminate the unitary matrix fuzzy to get each information bit Corresponding combination code section,and the complete combination code sequence can be achieved by combining with each part of the combined code.The simulation results show that,the method can be accurate estimation performance at low SNR.
出处
《科学技术与工程》
北大核心
2014年第22期45-51,共7页
Science Technology and Engineering
基金
国家自然科学基金项目(61371164
61071196
61102131)
信号与信息处理重庆市市级重点实验室建设项目(CSTC2009CA2003)
重庆市杰出青年基金项目(CSTC2011jjjq40002)
重庆市自然科学基金项目(CSTC2012JJA40008)
重庆市教育委员会科研项目(KJ120525
KJ130524)资助
