LFM信号检测算法的性能分析

分析了匹配傅里叶变换（MFT）、多通道数字去斜和多通道自相关三种检测线性调频信号（LFM）的算法，给出了其具体的实现流程。通过Matlab仿真验证了算法的可行性，给出了计算量和部分LFM信号参数估计的分析结果。分析表明，MFT算法性能最好，在多目标情况下实现对信号参数的估计；多通道自相关算法性能最差，但计算量小，而且算法流程便于工程实现。从工程应用的角度对算法作了一些改进，在应用中可以根据各算法的特点和实际需求选择合适的方法。

Performance Analysis of LFM Signal Detection Algorithms

Three LFM signal detection algorithms, matched Fourier transform, multi-channel digital deramping, and multi-channel autocorrelation, are analyzed. The detailed processing flows are given. Some simulations are done to validate these methods with Matlab. The computation amount and some parameter estimations for LFM signal are analyzed. It is demonstrated that the performance of MFT method is the best and the multi-channel autocorrelation detection is the worst. The former can estimate the parameters in multitarget situation. The latter can be realized easily and is of low computation amount. The methods are improved in the view of engineering applications. According to the characteristic of each method, the appropriate approach should be chosen to meet the needs in applications.