Research On Identification And Parameter Estimation Of Communication Jamming Signals

As the key technology in wireless communication countermeasure system,the detection and recognition technology of communication jamming signal plays an increasingly important role,which has an urgent need in practical application.Based on this research background,this paper takes the detection and recognition technology of communication jamming signals as the entry point for research,and analyzes and explores the key technologies of the detection algorithm,parameter estimation and pattern recognition of communication jamming signals.Firstly,this paper studies and analyzes two existing interference detection algorithms CME and FCME.Aiming at the problem that non-idle band interference detection is not considered in the existing scheme,an improved interference detection scheme suitable for non-idle band is designed.The interference detection is performed on the passband and the transition band of the signal by performing frequency domain block processing on the received signal.The simulation results show that under the same conditions,the proposed method exhibits better detection performance and wider application range than the existing methods.Secondly,this paper estimates the key parameters of the jamming signals,including the center frequency and the bandwidth.Aiming at the problem that the power spectrum of the jamming signals has many burrs and large fluctuations,the noise reduction and smoothing of the power spectrum are first adopted,and then the parameter estimation is performed,so that better estimation accuracy can be obtained.In this paper,based on the principle of EMD algorithm to remove high-frequency components and reserve low-frequency components for decomposition and reconstruction,the low-pass filter denoising method is used to perform filtering and noise reduction processing directly in the power spectrum domain to obtain a smooth power spectrum.Then the parameter estimation method based on frequency domain analysis is given.The simulation results show that the method adopted in this paper shows good estimation accuracy under low JNR,and the performance is stable and the complexity is low.Finally,for the pattern recognition method of communication jamming signals,this paper uses the traditional statistical pattern recognition method to classify and identify six typical communication jamming signals,including single-tone jamming,multi-tone jamming,linear sweep jamming,noise-modulated jamming,noise-AM jamming and noise-FM jamming.Firstly,the data preprocessing method is introduced.Secondly,a total of nine characteristic parameters of jamming signals are extracted in time domain and frequency domain respectively.And the variation trend of the characteristic parameters with the JNR is analyzed by simulation.At the same time,the algorithm principle of traditional decision tree and BP neural network is studied,and the identification scheme of jamming signals is given separately.In order to overcome the drawbacks of decision tree method,which is easily affected by threshold,high complexity of BP neural network and no theoretical basis for parameter setting,the combination of XGBoost algorithm and Bayesian optimization algorithm is adopted in this paper.The XGBoost algorithm is used to train the model and the Bayesian optimization is used to optimize the model parameters.Compared with the traditional parameters search method,the Bayesian optimization algorithm can find the global optimal parameter combination faster.The simulation shows that the proposed method has better recognition performance at low JNR,and it still has stable performance and low complexity on small data sets.