Research On Identification Friend Or Foe Signal Classification Technology Of Air Target

In modern warfare,the enemy and our fighters are constantly staggering in the airspace.It is necessary to quickly identify friend or foe for aerial targets that appeared in the combat area.The traditional Identification Friend or Foe(IFF)is divided into two types: cooperative and non-cooperative.The cooperative identification system uses inquiry or response to judge whether the air target is friend or foe,but if the target is not your own target,it cannot decode the friend or foe signal,so fast recognition cannot be achieved.In addition,due to electromagnetic interference in the battlefield environment,and the monitoring equipment will also malfunction,resulting in incomplete identification of the friend or foe signal received,which increases the difficulty of identifying and classifying the identification or foe signal.At the same time,there is a problem that it is difficult to effectively store the collected target signals and classification result information.In order to solve the above-mentioned problems,we study the related technologies for the classification of air target friend or friend identification signals,and explore methods that can quickly identify and classify the cooperative identification friend or friend signals.The main work of this article includes:(1)In view of the data missing in the received identification friend or foe signal data,this paper uses Singular Value Decomposition(SVD)to improve the traditional k-Nearest Neighbor(KNN)filling algorithm.SVD can get an approximate representation of the original data matrix,so the new filling algorithm can compensate The KNN filling algorithm ignores the defects of the inherent correlation between the data,and the filling effect of the data set with a larger missing rate is also better than the KNN filling algorithm.(2)For the cooperative IFF signal sent by an enemy target,although it cannot be decoded,it can obtain the pulse characteristics with timing changes in the signal waveform,and then use the machine learning algorithm to classify the obtained pulse characteristics.Prior information such as expert experience can realize rapid recognition of the identification friend or foe signal.This paper selects the pulse width of the identification friend or foe signal feature as the research object,and obtains the feature data set of the identification friend or foe signal by simulation.At the same time,it studies the decision tree classification algorithm,builds the decision tree classification model,and compares the decision tree model under different signal-to-noise ratio environments.Anti-noise ability is tested.The experimental results show that the decision tree model can complete the classification task of the identification of friend or foe with high recognition accuracy,and has strong anti-noise performance.(3)Finally,in view of the classification results and some other information about the air target,this paper designs a target signal feature library to effectively store it,completes the conceptual design,logical design and physical design of the target signal feature library,and finally the construction of feature database is implemented in My SQL database.