In this paper, a dual model unscented Kalman filter chaotic fitting breaking method is proposed to break chaotic direct sequence spread spectrum communication systems in the cases of low spreading factor, low signal-to-noise ratio or severe multipath fading. Based on the characteristic that the range of information symbol is a finite set, the proposed algorithm fits the original chaotic signal through different filters which work in parallel. The fitting errors are used to choose the optimum matching filter, thus to estimate the information symbols. Furthermore, an error-controlling-factor is introduced to increase the distance of model based tracking errors, which can not only facilitate the information extracting process, but also reduce the influence of noise and multipath fading. Theoretical analysis and simulation results prove that the proposed algorithm is superior to the existing breaking method.%针对已有混沌直接序列扩频通信系统的破译算法在低扩频因子和较大多径衰落下无法有效破译信息码的问题，提出了一种基于双模型无迹卡尔曼滤波混沌拟合的破译算法．所提算法联合多模型滤波原理，利用信息符号为有限集的特点，针对其不同取值分别建立对应的滤波模型．各模型下滤波器并行工作，利用广义同步系统同时拟合原混沌系统并估计混沌直扩信号，通过估计误差确定最佳匹配滤波模型，从而得到信息符号的估计．进一步通过引入误差控制因子，增大了不同模型下估计误差的距离，不仅有利于信息符号判定，并且减小了噪声和多径衰落对破译结果的影响．理论推导和仿真结果均证明提出的算法优于已有破译算法．
展开▼
