An Efficient Method for Parameter Estimation and Separation of Multi-Component LFM Signals

Shaohang Liu,Zhenkun Lu,Ji Qiu,Qinghua Huang,Cui Yang
DOI: https://doi.org/10.2139/ssrn.4134195
2022-01-01
SSRN Electronic Journal
Abstract:This paper proposes an efficient fast parameter estimation and separation algorithm for multicomponentlinear frequency modulation signals based on the fourth-order origin moment of fractionalspectrum and fractional domain filtering. Aiming at the problems of low estimation accuracyand large computational complexity of traditional two-dimensional search and time-frequencyanalysis methods, the algorithm applies an efficient fractional time-frequency analysis methodto determine the initial interval and search center of the optimal rotation order. Moreover, thealgorithm proposed employs the good impulse performance and anti-noise performance of thefourth-order origin moment of fractional spectrum to formulate a reasonable search strategy of theoptimal order, which reduces the amount of calculation while ensuring the accuracy of the parameters.A step-by-step separation and filtering technique for signals in the fractional Fourier domainis further implemented for the problem of interference between multi-component signals, whichsuppresses the influence of strong components on the parameter estimation of weak components.The error analysis of the linear frequency modulation signal is carried out, and the Cramer-RaoLower Bound for the unbiased estimation of the parameters is derived. The final simulation resultsprove the accuracy and effectiveness of the method in noisy environments.
What problem does this paper attempt to address?