Radar Individual Identification And Jamming Decision

Electronic countermeasures are designed to disrupt the search,tracking,and identification of enemy radars to protect the safety of our equipment.Electronic countermeasures mainly have the following two stages:One is electronic reconnaissance.Electronic reconnaissance is the prerequisite of electronic countermeasures.It analyzes the technical parameters of the target radar signal through the target radar signal intercepted by the reconnaissance aircraft,and then obtains the relevant information of the launching equipment and working status of target radar through technical means;The other is electronic interference.Electronic interference is when the jammer emits electromagnetic signals to interfere with the target radar,so that the enemy radar equipment cannot resolve the true echo signal or obtain the false signal emitted by the interference equipment.This paper will study the problems in the three aspects of individual radiator identification in the electronic reconnaissance stage,working status identification,and interference decision-making in the electronic jamming stage.Radar radiation source transmitting equipment has a "fingerprint feature" that is different from other devices due to the difference in hardware,so the individual radar radiation source can be identified through the difference of "fingerprint feature".The radar radiation source signal obtained by the reconnaissance aircraft may be distorted and the matching sequence is relatively high when the shape matching algorithm is used.This causes the radiation source signal in the template library to be unable to obtain the target sequence by summing and averaging.In response to this problem,this paper uses the dynamic time warp barycenter averaging algorithm(DTW Barycenter Averaging,DBA)to obtain the target sequence quickly and accurately.Aiming at the problem of individual radiation source identification,a dynamic time warping(Adaptation Weight Function,awf-DTW)algorithm suitable for the adaptive weight function of the radar radiation source signal is proposed to solve the negative ranging effect caused by the bending and twisting fingerprint features.After the reconnaissance aircraft obtains the technical parameters of the target radar radiation source through technical analysis,it can recognize the working status of the radar.The reconnaissance aircraft continuously obtains radar radiation source signals,and the characteristic of the radiation source signal data is that the samples are not balanced,and the data with few samples is highly threatening.Aiming at the problem of possible concept drift and imbalance of sample categories in the massive radiation source data,this paper proposes an adaptive windowing detection and diversity based ensemble learning algorithm(Adaptive Windowing Detection and Diversity Based Ensemble Learning,ADEL),the algorithm can adaptively solve the concept drift phenomenon in the data stream and does not need to retrain all the data,thereby speeding up the model training time and achieving a more complete comprehensive index.Radar interference decision-making requires the jammer to change the interference pattern and the allocation of interference resources in real time according to the working status of the enemy radar to protect our equipment,therefore,radar jammers need to make accurate and fast decisions.This paper proposes an adaptive Q-Learning jamming decision algorithm.The algorithm divides the different working states of the enemy radar into different threat levels,and adjusts the jamming decision of the jammer according to the change of the threat level.The training results adaptively adjust the hyper parameters to accelerate the model convergence,and on the basis of ensuring the correct interference path,it has a shorter learning time than Q-Learning with fixed hyper parameters.