一种具有窄带干扰抑制能力的改进型LSL插值算法  

The performance of narrow-band interference suppression in DSSS using improved LSL interpolators

作　　者：孙志国[1] 郭黎利[2] 顾学迈[1] 

机构地区：[1]哈尔滨工业大学通信技术研究所,黑龙江哈尔滨150001 [2]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《电路与系统学报》2008年第3期13-17,12,共6页Journal of Circuits and Systems

基　　金：高等学校博士学科点专项科研基金资助课题(20050217016);哈尔滨工程大学基础基金资助课题

摘　　要：本文提出了一种基于QR分解技术的最小二乘格形插值算法—QRD-LSL插值算法来抑制直扩系统中的窄带干扰。该算法利用直扩信号、噪声信号和窄带干扰信号数据样值间相关特性的差异,通过对直扩系统接收端数据样值序列进行插值运算来估计窄带干扰信号成分,从而达到抑制直扩系统中窄带干扰的目的。由于该算法引入了QR分解过程避免了对数据矩阵求逆运算,并引用动态范围较小的新变量参与阶递推运算,极大程度上提升了算法在稳态特征、收敛速度和计算复杂度等方面的性能。仿真结果表明:QRD-LSL插值算法较LSL插值算法和LSL预测算法至少可获得4dB的处理增益,且计算量与滤波阶数呈线性关系。An improved QRD-LSL interpolation algorithm based on QR-Decomposition is proposed for Narrow-Band-Interference （NBI） suppression in DSSS. This algorithm utilizes the discrepancy in the predictability between NBI and the spread spectrum signal, and then uses a linear interpolation of the received signal to estimate NBI. Because of QR-Decomposition operation instead of inverting the correlation matrix and a novel variable with a small dynamic range, QRD-LSL interpolators outperform LSL predictors and LSL interpolators. It is shown that the QRD-LSL interpolators can obtain more 4dB processing gain than LSL interpolators or LSL predictors, and its total computational complexity scales linearly with the interpolation order.

关 键 词：QRD—LSL插值算法 QR分解 窄带干扰 直扩系统 

分 类 号：TN914.42[电子电信—通信与信息系统]