基于稀疏贝叶斯重构的多跳频信号参数估计

Parameter estimation for multiple frequency-hopping signals based on sparse Bayesian reconstruction

针对宽带条件下多跳频信号的参数估计问题,利用多跳频信号在空间频率域上的稀疏性,提出了基于稀疏贝叶斯重构的空间频率估计方法.通过重构信号获得信号瞬时频率估计,在此基础上完成了波达方向信息估计.为了提高低信噪比条件下参数的估计性能,采用形态学滤波的方法对得到的时频图进行修正,在修正的时频图上完成了信号频率集和跳周期的精准估计.仿真实验表明:该算法在信噪比低于0dB的情况下仍能够取得良好的估计性能.

With the purpose of estimating parameters and signal sorting in wideband frequencyhopping networks,a method for estimating spatial frequencies based on sparse Bayesian reconstruction was proposed using the sparsity of multiple hopping signals in spatial frequency domain.First,the instantaneous frequencies were obtained by signal reconstruction,based on which the direction of arrival（DOA）were estimated,and the signals could be selected consequently.To conduct parameter estimation in low signal-to-noise ratio（SNR）conditions,the time-frequency spectrum was calibrated in morphological filtering methods,which allowed for exact estimation of signal frequency sets and hopping periods.Simulation results show that excellent estimation performance of proposed algorithm could be obtained even when the SNR is lower than 0dB.