Signal Recognition And Parameter Estimation Based On FRFT

Electronic reconnaissance is an important component of electronic countermeasure, and accurate radar signal identification is the key task in the electronic reconnaissance system. In recent years, the electromagnetic environment becomes more complex, it is difficult to obtain effective identification results if only rely on the traditional method and the traditional parameters. How to timely and effectively detect radar signal is a major research direction in the field of electronic countermeasures. Fractional Fourier transform is a new time-frequency analysis method which has a free parameter transform order p , with improvement of its relevant theory and numerical algorithm, it is gradually used in the field of non-stationary signal processing while radar signal is a typical non-stationary signal. This paper mainly studies signal recognition and parameter estimation based on the Fractional Fourier transform which can be divided into three parts:Firstly, modulation mode recognition of several typical modulation signal. A joint recognition algorithm is proposed based on the reduced Fractional Fourier transform and the short time Fourier transform, which combines the idea of decision tree to make full use of respective advantages of Fractional Fourier transform and Short-time Fourier transform, it can reduce the amount of calculation and improve anti-noise property. An improved algorithm is proposed based on fractional frequency spectrum fourth order origin moment,which uses three characteristics of envelope curve include the curve peak value, order p and kurtosis to construct a feature vector, and then dynamic clustering method is used to identify the signal, which can achieve a good separation between the class and aggregation within the class, and these three characteristics are effective complement for traditional characteristic parameters.Secondly, radar signal detection and parameter estimation. An improved linear frequency modulation signal detection and parameter estimation algorithm is proposed based on the fast Radon-Wigner transform and the Fractional Fourier transform, by using the fast Radon-Wigner transform to rough estimate modulation frequency in advance, then only needs to search in a small fractional domain and avoid global search, which can reduce the amount of calculation. An improved linear frequency modulation continuous wave signal detection and parameter estimation algorithm is proposed based on the periodical fractional order spectrum fourth-order moments, to simplify the Fractional Fourier transform can reduce amount of calculation under the premise of guarantee calculation precision, at the same time the fractional order fourth-order moments can improve anti-noise performance, the simulation shows that estimation precision of the improved algorithm is well in low SNR environment.Finally, the local frequency spectrum information is more important for the multi-component signal processing,and the Short-time Fractional Fourier transform is a good method. Through analyzing influence factors of the time-frequency resolution of the Short-time Fractional Fourier transform, we found it is greatly influenced by the width of window function and the search interval. For the multi-component linear frequency modulation signal detection and parameter estimation, an improved algorithm is proposed based on the idea of "CLEAN", by adjusting the width of window function to improve the time-frequency resolution, according to the size of signal component energy to extract signal component in order, and by using the secondary search method to search the optimal order domain which can reduce the amount of calculation,the simulation verifies the effectiveness of the improved algorithm.