Research On Modulation Recognition And Parameter Estimation Of Pulse Compression Radar

The emergence of low probability radar has improved the radar's survival rate and antijamming ability in the battlefield environment,and pulse compression radar,as an important implementation form of low probability radar,has been widely used in new radar systems.Therefore,the research on characteristic analysis,recognition and parameter extraction of pulse pressure radar has become a hot spot in related fields.This article focuses on common pulse compression radar styles,and launches the research of reconnaissance processing technology.The full text can be divided into three parts:1.Introduced the mathematical model of typical modulation signals(chirp signal,dual chirp signal,sine FM signal,even second FM signal,phase-encoded signal)and analyzed the time-frequency characteristics of each modulation signal,as the following text The classification and recognition provides materials for dividing the sample space;introduces classic time-frequency analysis tools,including short-time Fourier transform and wavelet transform;finally,introduces common signal preprocessing methods: short-term filtering and wavelet denoising.2.Two classical modulation recognition algorithms are studied and improved on the basis of them.The improved algorithm mainly realizes the classification of the above five kinds of signals.First of all,in view of the difficulty in extracting the time-frequency curve with the traditional algorithm,it is proposed to combine the I-Rife sine wave frequency estimation algorithm with the short-time Fourier transform,which greatly improves the time-frequency curve recovery effect.The simulation proves that under the condition of SNR>2d B Can realize signal modulation recognition.In addition,the traditional instantaneous autocorrelation spectrum method has been improved,and short-term filtering and noise reduction is added before peak statistics,which makes the autocorrelation spectrum feature more obvious,and the system's anti-noise performance is further improved.Achieve a 90% recognition rate.3.According to the characteristics of each modulation signal,several classic parameter estimation schemes are studied,two of which are improved and simulations have verified the superiority of the improved scheme.First,the traditional fractional Fourier transform and autocorrelation spectrum analysis method are combined and used to estimate the parameters of chirp signals.The improved algorithm performs a second fine search on the basis of the rough estimation value obtained by the autocorrelation spectrum method,which has both good estimation accuracy and fast analysis speed.Second,improvements are made to the shortcomings of poor real-time performance and unstable estimation performance when the cubic phase function estimates even second-order FM signals.The improved scheme reduces the amount of calculation by combining fast Fourier transform,and uses all samples in the calculation process to improve the stability of the algorithm.In this paper,from the perspective of reducing the amount of calculation and improving real-time performance,the existing parameter estimation algorithm is improved,so that the algorithm can be smoothly applied in engineering practice.