DOA Estimation For Wideband Chirp Signal Using FRFT

Array signal processing is an important field in signal processing. And non-stationarywide-band array signal processing is the emphasis and difficult to study in this field. So itis important and worthy to study the theory and method for non-stationary wide-bandsignal processing. The main work of this paper is aimed on this subject. It uses FractionalFourier Transform(FRFT) to process the wide-band signal such as chirp signal. FRFT canreduce the signal band width and transform a wide-band signal to likely narrow-bandsignal and then most traditional DOA algorithm for narrow-band signal can be used on thetransform result to estimate the arriving direction of signal source. The paper proposessome algorithm combined with FRFT and traditional algorithm, explains the algorithmsand gives simulating result of these algorithm.(1)Wide-band DOA algorithm based on WLCFWide-band linear chirp FRFT method (WLCF) algorithm canestimate arrivingdirection with sensory array and chirp signal. In time-frequencyplane, chirp signal can bedescribed as a line that has a angle with time axes.Use the rotational character of FRFT,Transforming the chirp signal with FRFTwith angle that is vertical to the chirp signalin time-frequency plane, the power of signal is focused in transformresult in a narrow field.The proposed algorithm uses the relation between chirpsignal and time-frequency plane,focuses the power in a new time-frequency plane and improves the signal-noise rate(SNR). Estimate the arriving direction with traditional DOA method for narrow-bandsignal, it can improve the accuracy in some degrees.(2) FRFT-based algorithm for wideband DOA estimationUse the time-frequency plane characteristic of fractional Fourier transform rotation,the time-frequency distribution of the direction vector can be approximated by fittingparallel to the horizontal axis at the new time-frequency plane, and thus achieve a narrowbandwidth, so this can achieve a narrow bandwidth, see the wideband array signal as thenarrow-band array signal to solve the problem of direction vector varies with frequency inDOA estimated,which is the advantage of this algorithm. Compared with the classicwideband DOA estimation algorithm methods, this algorithm method have no need of thepre-estimate of the angle and center frequency, reduce the error caused by the“Focus”after the segment,so this algorithm has higher estimation accuracy. (3) FRFT-based algorithm for wideband DOA estimationUsing the time-frequency plane characteristic of fractional Fourier transform rotation,the time-frequency distribution of the direction vector can be approximated by fittingparallel to the horizontal axis at the new time-frequency plane, and thus achieve a narrowbandwidth, so this can achieve a narrow bandwidth, see the wideband array signal as thenarrow-band array signal to solve the problem of direction vector varies with frequency inDOA estimated,which is the advantage of this algorithm. Compared with the classicwideband DOA estimation algorithm methods, this algorithm method have no need of thepre-estimate of the angle and center frequency, reduce the error caused by the“Focus”after the segment,so this algorithm has higher estimation accuracy.(4) Wideband DOA estimation based on Fractional-CumulantIn the scheme of wideband DOA rotational Cumulant estimation based on fractionalFourier transform for Chirp signal, this method is combined fractional cumulant algorithmwith high order cumulant. The algorithm has rotation characteristics in time frequencyplane and is not sensitive to the Gauss process, and is used in wideband DOA estimation.We can see that this algorithm is not only to reduce the bandwidth, but also can effectivelyrestrain Gauss noise, and the simulation verified the effectiveness of our algorithm.In summary, the wideband DOA estimation is studied by rotational characteristics ofFRFT, and got some progress. The proposed algorithm, which is based on shorting thebandwidth, is different form the common idea of processing the wideband DOAestimation. This is outstanding point of our work. Moreover, the advantage of thisalgorithm can prevent from losing the message and reduce the estimation accuracy.Compared with the narrowband DOA estimation, these algorithms perform better andhave higher accuracy.