Study On The Feature Analysis Of The Lenear Frequency Modulated Signal Based On The Fractional Fourier Transform

The extraction of radar signals’ feature is one of the important methods toidentify active targets in modern electronic warfare, and the characteristics of thesignal will provide a basis for signal sorting. Linear frequency modulated signal isthe most widely used signal in the radar and other electronic systems, so theparameter estimation of the LFM signal is of great significance. The subject of therequirements for engineering applications, according to the characteristics of theLFM signal, the paper mainly studied the fractional Fourier transform to estimatethe signal parameters.Firstly, the paper studies the requirements of the continuous Fractional Fouriertransform discretization. In order to select an algorithm which is suitable forengineering application, this paper chooses the feature decomposition algorithmthrough the analysis of several types of discrete algorithms. And conduct a detailedtheoretical and simulation analysis of the approximation between the discretealgorithm and continuous fractional Fourier transform algorithm.Secondly, in order to extract the feature of the linear frequency modulatedradar pulse signal, this paper derives the mathematical expression of the signalparameters theoretically, analyzes the impact of sampling frequency and noise onparameter estimation, proposes a method to reduce the computational quantity. Toverify the correctness of the theory this paper uses the matlab simulations.Finally, it studies the implementation of the discrete fractional Fouriertransform algorithm on FPGA. Propose a FPGA structure of the algorithm andanalyze the parallelism of the algorithm and utilize the CORDIC algorithm torealize the calculation of the eigenvalue vector. To implement sequential control ofeach sub-module it uses the synchronous state machine. This paper further analyzesthe works of the algorithm computation time and resource consumption and thereasons that caused the error. Through calculating the square wave signal in theFPGA and matlab, we get the conclusion that operating results in FPGA matches thematlab.