Uwb Lfm Signal Detection And Parameter Estimation Methods

Ultra-wide band (UWB) radar is one of the important developing aspects ofmodern radar technology because of its ability in super-resolution andanti-interference. How to detect the UWB radar signal and estimate the signalparameters in low signal to noise ratio condition is very important. Many newUWB radar signal detection and parameter estimation algorithms have beendeeply and widely studied in some degrees in this dissertation. The mainconcerned signals are Linear Frequency Modulated (LFM) signal.This paper delves into the research of signal detection, parameterestimation algorithms and the effects caused by the parameter's imprecision. Themain work can be summarized as follows:1. On the condition of ultra-wide band, it's necessary that makereconnaissance of signal in the negative signal noise ratio (SNR). Author studiesthe time-frequency presentation (TFR) including the linear short-time Fouriertransform (STFT) and the quadratic Wigner-Ville and Radon-Wigner etc. We candraw the conclusion that the resolution of STFT relatively poor, tomono-component signal WVD is fairly good but the cross-terms make theperformance worse seriously. Radon-Wigner overcomes the disadvantages buthas large computation amount.2.Make wavelet transform to the signal and analyze the relation ofamplitude of wavelet coefficient and scale. LFM in frequency domainapproximate to a rectangular. According to the different singular exponentialbetween signal and noise, we can detect the signal for the different changeabletendency. By the way of de-chirp can improve the SNR .In combination withwavelet transform the performance of detection is better.3. To overcome the disadvantages of typical TFR analyze the matchFourier transform (MFT) and optimize algorithm. The results of severalexperimental tests conducted both on mono-component and multi-componentsignals are presented; advantages over other solutions are obvious.