基于能量分离算法的多分量线性调频信号频率估计

Frequency estimation of multicomponent LFM signals based on energy separation algorithm

对基于能量算子的能量分离算法进行了改进,提高了该方法用于多分量线性调频信号瞬时频率估值的抗噪性.采用平滑样条插值的方法对离散信号进行插值,然后再对其应用连续信号的能量分离算法.仿真表明,改进后的算法在噪声环境下,比直接利用对应离散信号的分离算法估值更准确,当信噪比从5 dB以5 dB为步进,增大至30 dB时,平均估计误差约减小10%.

Improved present energy separation algorithm that is based on energy operator. The noise immunity was improved when it was applier to estimate the instantaneous frequency of multicomponent LFM signals. Smooth spline interpolation was used on discrete signals, and then continuous algorithm was used to estimate the frequencies of muhicomponent LFM signals. Simulated results show that the proposed method has better performance than directly applying the discrete separation algorithm to discrete signals. And the estimating error of the proposed method is 10% lesser under SNR from 5 dB to 30 dB stepped by 5 dB.