Research On Radar Signal Detection And Parameter Estimation Technology In Complex Electromagnetic Environment

In recent years,electronic warfare has become increasingly important in modern warfare,addinga new domain for battle.In the complex electromagnetic environment,radar signal electronic reconnaissance is the forerunner of electronic warfare.The method is to intercept the unknown radar radiation source signal and analyze its related characteristic parameters.Under this premise,this thesis firstly detects the received signal to determine whether it contains a radar target.Secondly,when it is judged to be a valid radar target,the modulation method of the radar signal is identified to determine the modulation parameters of the signal.Finally,the incoming wave direction angle estimation is performed on the signal to determine the position parameter of the signal.The specific research contents are as follows:(1)An adaptive constant false alarm detection algorithm is studied for radar target signal detection.In view of the lack of information priori characteristic of noncooperative parties in radar signal reconnaissance.Firstly,the non-coherent accumulation technique based on short-time Fourier transform is used to obtain the accumulation gain to improve the detection performance of the target signal.Secondly,improved the degradation of cell averaging constant false alarm rate in multi-objective detection.Simulations demonstrate the effectiveness of this method.(2)The recognition of radar signal modulation mode based on two-way parallel neural network is studied.After the signal characteristics are transformed into amplitude and phase characteristics,they are used as the input of neural network.The parallel network model of long-term and short-term memory network optimized by Resnet and attention mechanism is designed.Combine the advantages of the upper and lower branch networks.Extract the spatial and temporal features of the signal separately,and then perform network output.The simulation proves that the modulation methods of the six radar signals mentioned in the paper can be well identified under the lower signal-to-noise ratio.(3)For the direction of arrival(DOA)estimation of radar signals,an extreme learning machine(ELM)and spatial smoothing algorithm is proposed to solve the DOA estimation of coherent signals.At the same time,seagull algorithm was used to optimize ELM model parameters and improve the effect of DOA estimation.First,the signal is spatially smoothed for decoherence.Then the upper triangular matrix of the signal sample covariance matrix is used as the input of the model,and the source angle is the output of the model.Finally,the trained model is used to estimate the signal angle,simulation results show that the algorithm has better direction finding accuracy and performance.