基于粒子滤波的单通道正弦调频混合信号分离与参数估计

发布时间：2019-03-27 16:38

【摘要】：以单通道正弦调频(SFM)混合信号为研究对象,提出了基于粒子滤波的正弦调频混合信号分离与参数提取方法.针对正弦调频混合信号频率无跳变的特征,提出了一种基于粒子滤波的相位差解混叠算法,并通过源信号相位差解决了本算法中粒子滤波高维状态空间降维问题,提出了一种适合高维状态空间的似然函数模型,比较固定长度粒子估计值和真实值误差,进而准确衡量粒子权重.通过在重采样后引入MCMC转移,解决了静止参数下粒子多样性降低问题,有效提高粒子滤波迭代收敛速度.从而在先验知识仅已知信号调制方式的情况下,完成对单通道正弦调频混合信号的参数提取,并通过重构信号完成正弦调频混合信号分离.最后通过仿真分析发现,该方法能够有效的实现正弦调频混合信号的分离与参数估计.
[Abstract]:In this paper, a single-channel sinusoidal frequency modulation (SFM) mixed signal is used as a research object, and a method for separating and extracting a sinusoidal frequency-modulated mixed signal based on particle filtering is proposed. A phase difference de-aliasing algorithm based on particle filtering is proposed for the frequency-free characteristic of the sinusoidal frequency-modulated mixed signal, and the dimension reduction of the high-dimensional state space of the particle filter in the algorithm is solved by the phase difference of the source signal. A kind of likelihood function model suitable for high-dimensional state space is proposed, and the fixed-length particle estimation value and the real-value error are compared, and the particle weight is accurately measured. By introducing the MCMC transfer after the re-sampling, the problem of the reduction of the particle diversity under the static parameters is solved, and the convergence speed of the particle filtering iteration is effectively improved. So that the parameter extraction of the single-channel sinusoidal frequency-modulation mixed signal is completed in the case of prior knowledge only the signal modulation mode is known, and the sinusoidal frequency-modulation mixed signal is separated by the reconstruction signal. Finally, through the simulation analysis, it is found that the method can effectively realize the separation and parameter estimation of the sinusoidal frequency-modulated mixed signal.
【作者单位】：南京理工大学电子工程与光电技术学院;
【基金】：国家自然科学基金(批准号:61301216,61171168,60702016) 省部委基金资助的课题~~
【分类号】：TN911.7

【相似文献】