论文部分内容阅读
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.
Taking single-channel sinusoidal frequency modulation (SFM) mixed signal as the research object, a new method based on particle filter is proposed to separate and extract sinusoidal frequency modulation (FM) mixed signal.Aiming at the characteristic of no-hopping of the frequency of sinusoidal FM signal, Phase difference and anti-aliasing algorithm. The phase-difference of the source signal is used to solve the problem of dimensionality reduction in the high-dimensional state space of the particle filter in this algorithm. A likelihood function model suitable for the high-dimensional state space is proposed. Real value error, and then accurately measure the weight of particles.It can solve the problem of particle diversity reduction under static parameters and effectively improve the iteration convergence rate of particle filter by introducing MCMC transfer after resampling.Therefore, only the known signal modulation method In this case, the parameters of the single-channel sinusoidal FM signal are extracted and the sinusoidal FM signal is separated by the reconstructed signal.Finally, the simulation results show that this method can effectively separate and parameterize the sinusoidal FM signal.