| If the phase of signal varies linearly with time, the complex signals combined withthis kind of signal can be ideally analyzed using Fourier Transform, for the phase oforthogonal base function varied linear with time and so get to match with the variationalrule of signal phase. But, in the natural and artificial signal, a lot of signal phases don'tvary linearly with time, for example, we usually use chirp signal in the radar technology,frequency step signal, linear modulation component of the ground echo in the SARtechnology, linear modulation component of the moving target echo, linear modulationcomponent while the airborn radar detecting the moving target, etc. The result ofanalyzing these signal with Fourier Transform is energy distribution at the differentfrequencies, it is to say, the projection of signal is on base function. To this kind ofsignal, Fourier Transform can not be dissatisfied with our need, so many scholars putforward different methods to resolve the problem according to the application inpractice, but they have not got ideal method still and been researching and delvingdeeply now. Analyzing the radar echo model, we shall understand that no matter whether theground-base radar or airborne radar or space-base detects the diversity targets, signalphases varied with time high order, so processing signal varied with time high order isone of necessarily resolved problems in modern radar. Depending on the base function of the Fourier Transform matching the signalphase varied linearly with time, so it can provide a useful idea for processing the phasechanged nonlinearly with time and unknown parameter signal. To process this kind ofsignal, we can design a transform to make the base function phase varied with timematching the signal phase, so it may be used to process the signal phase variednon-linear with time. This method is called as Matched Fourier Transform. ·The keystone of Matched Fourier Transform is the rule of base function phasevaried with time is the same as the rule of signal phase varied with time, that is calledmatch, but existed a restricted condition, if the function varied with time is ξ(t) , thederivative of ξ(t) is continuous, and d ξ(t) ≥ 0 or dξ(t) ≤ 0 is consistent in [0,T], dt dtξ(0) = 0. The orthogonal character of Matched Fourier Transform base function andMatched Fourier Transform pairs existing are proved in this paper. If a signal is chirp,the spectrum by Fourier Transform is within specific bandwidth. Every spectrum line Key Lab for Radar Signal Processing　 ABSTRACT　　 IV　share a part of signal energy. When there is noise in signal, it is disadvantage to detecttarget signal. When there are multiple targets in the signal, maybe the targets signalsspectrum will overlap. It is very difficult to distinguish the targets number. Here toestimate parameter is not completely accurate. However, the signal processed byMatched Fourier Transform, every chirp signal can form only a spectrum line. Becausethe Matched Fourier Transform can make the chirp signal energy centralized on thespectrum line and the same time the noise spectrum lies howbeit in a specific bandwidth,it is very advantage to detect the chirp signal. It is the similar to the detection ofDoppler signal processed by the Fourier Transform. For a target signalcan form only aspectrum line, it is very easy for the parameter estimation and the resolution of thetarget signal. In order to adapt to modern digital processing technology, it is alsodiscussed in this paper the discrete Matched Fourier Transform, sample frequency andsample number effect the Matched Fourier Transform, the relationship betweenperiodical signal sample frequency. The result shows some special characters ofMatched...
|
PDF Full Text Request