| As a kind of time-varying signals, Chirp signal widely appears in systems such as communication, radar, sonar, biomedicine and earthquake exploring system and also be used as signal model for a lot of natural phenomena. The detection, the parameters including initiative frequency, frequency modulation ratio and the direction of arrival estimation and the time-frequency representation for Chirp signal were researched in this dissertation, using the Radon-Wigner transform, Smoothed Pseudo Wigner-Ville Distribution (SPWVD) with Radon transform, reassignment time-frequency distribution with Radon transform, and fractional Fourier transform (FRFT) methods.The time-frequency distribution performance of WVD, Pseudo Wigner-Ville Distribution (PWVD) and SPWVD for the multi-component Chirps was described from the time-frequency concentration, the cross terms reduction, and the discriminability. Though WVD is a better time-frequency representation for only one Chirp component, the interferences of cross terms of the bilinear time-frequency distributions occurred for the multi-component Chirps. PWVD and SPWVD Using window function smooth can restrain the cross terms, but it results in a deterioration of the time-frequency concentration and the discriminability. Therefore, WVD, PWVD and SPWVD are not the perfect as a time-frequency analysis tool for the multi-component Chirps.Because the presence of the multi-component Chirps is barely identified directly from time-frequency distribution due to the existence of the cross terms and noise, the method of the detection and the basic parameters estimation for the multi-component Chirps was proposed. This method is based on bilinear time-frequency distribution and Radon transform, which regards the bilinear time-frequency distribution of the signal as a beeline, applies the principle of the integral along beeline of Radon transform and finds the local maximal and relevant coordinate in order to detect the signal and estimate its basic parameter. The cross terms of WVD come into being fake peaks by Radon transform. The time-frequency concentration will fall when the cross terms are smoothed by window function. To resolve this question, this dissertation put forward the improved method of the reassignment time-frequency with Radon transform. The proposed method gives attention to restraining the cross terms and increasing time-frequency concentration, it can not only eliminate false detection resulted from fake peak, but also increase the accuracy of the detection and parameters estimation and reduce the interference of noise. The corresponding simulation experiments were given and the simulation results shown that the proposed methods are effective.The fractional Fourier transform(FRFT) interprets signals as the combination of the orthonormal linear frequency modulation (LFM) primary functions. There will be conglomeration of energy in fractional Fourier domain when the fractional Fourier domain accord with the frequency modulation ratio of the Chirp component. On condition that the length of signal is infinite, the conglomeration of energy will represent as an impulse function, and it will represent a local maximal amplitude when the length of the signal is finite. Through FRFT Gauss white noise is still gauss white noise. Therefore there is an appropriate angle to get right fractional Fourier domain, in which the Chirp component unfold as conglomeration of energy, while gauss white noise can not unfold as conglomeration of energy in any fractional Fourier domain because it equably distribute on the whole time-frequency plane. For Chirp signals FRFT is a kind of suited transform, Different suited angles can separate the different Chirp component when their frequency modulation ratios are different. With this characteristic, a method for the detection and parameter estimation of Chirp signals based on FRFT. The simulation experiments were given and the simulation results shown that the proposed method can well and truly detect Chirp component and can accurately estimate every initiative frequency, frequency modulation ratio and amplitude. Chirp signal is widely applied in systems such as communication, radar, sonar, and so on. The direction of arrival is an interesting parameter. Because Chirp signal is wideband, its direction vector is time-varying, so it is difficult to process Chirp signal, besides Chirp signal is a typical non-stationary signal, routine subspace methods are not appropriate. The array signal processing is an important branch of modern signal processing, whose essence is using the spatial disperser arrangement sensor array and the multi-channel receiver to gain multi-dimensional information such as the signal's information in time domain and airspace , and so on. The final goal is to detect the signal and pick the parameters of the signal. Time-frequency analysis is an effective method of processing non-stationary signals. It transforms the unidimensional time signal to two-dimensional time-frequency domain, which reveals the change tendency of each frequency component along with the time in the signal. Therefore time-frequency analysis and array processing can be combined to enhance the performance of the Chirp signal position estimate. Wigner-Ville distribution has very good time-frequency accumulation performance for Chirp signals, the signals can be separated in the time frequency plane, and each signal source can be processed alone. Then the disturbance between the signals can be eliminated. Considering realization, the method based on PWVD and MUSIC was proposed to estimate DOA of Chirp signals. But when two signals in the very much near situation, the cross-terms disturbance between the signals is serious, it is difficult to sort the time-frequency point. Then the method based on reassignment SPWVD and MUSIC was proposed to estimate DOA of Chirp signals. Then, disturbance of cross-terms was eliminated, the precision of DOA estimate is elevated in the low signal-to-noise ratio situation, and anti-noise performance of the algorithm was enhanced. The simulation experiments were given and the feasibility and validity of the corresponding algorithms had been confirmed. There is an energy accumulation in the suitable fractional Fourier domain of aChirp component. Fractional Fourier transform reveals a very strong ability to select signals regarding the Chirp signals. If the correlation matrix for Chirp DOA estimate is constructed in fractional Fourier domain directly, there is no disturbance from cross-terms, the anti-jamming performance of the estimate algorithm is enhanced. The correlation matrix was constructed in fractional Fourier domain using fractional Fourier transform of Chirp signals, the DOA estimate of Chirp signals were carried out with MUSIC algorithm. The disturbance from cross terms disappeared, the difficulty to sort the time-frequency point from bilinear time-frequency distribution was avoided, and each Chirp source can be separated conveniently. Simulation experiments were given, and the simulation results show that the algorithm is feasible and valid. |
PDF Full Text Request