Research On Operation Mechanism Of Radar Waveform Library And Signal Recognition Method

In recent years,with the continuous development of radar technology,the electromagnetic environment on the battlefield is becoming more and more complex.The traditional radar signal recognition algorithm is facing new challenges.Firstly,with the increase of the number of radar signals and feature dimensions,the amount of data used for signal recognition continues to increase.It is necessary to store and manage a large amount of data efficiently and reasonably.Secondly,in practice,the new types of signals are appearing,and the old types of signals are eliminating,while the existing signal recognition algorithms can not effectively learn this change,resulting in the reduction of the accuracy and efficiency of signal recognition.To solve these problems,establishing a dynamically managed radar waveform library operation mechanism and combining it with the signal recognition method is an effective means to improve the performance of signal recognition.The core includes the establishment of operation mechanisms such as sample implantation,memory and forgetting,partition recognition and scheduling.Therefore,this paper mainly carries out the research in the following aspects:(1)Aiming at the problems of structural redundancy,fixed samples and difficult expansion in the storage of a large number of signal waveforms and characteristics,a radar waveform library with dynamic data management is constructed.By designing the data storage structure of radar waveform library and introducing intelligent operation mechanisms such as sample implantation,memory and forgetting,partition identification and scheduling,the efficient storage and dynamic organization and management of signal samples are realized.(2)Aiming at the problem that it is difficult to accurately recognize new types of samples in the existing signal recognition system,a radar signal incremental recognition method based on sample screening strategy is proposed.Through the clustering screening and incremental learning implantation of unknown class radar signals,the expansion of new class signal samples is realized,and the recognition accuracy of new class samples is effectively improved.(3)Aiming at the problems of high computational resource overhead and many redundant categories in signal recognition system,a radar signal reduction recognition method based on dynamic sample forgetting is proposed.By introducing the memory model to evaluate the importance of radar signals,combined with the reduction learning forgetting process,the non-important signal category samples are eliminated,and the optimal balance between recognition accuracy and efficiency is realized.The methods proposed in this paper have been verified by simulations.The results show that these methods can recognize radar signals efficiently and accurately in the case of iterative updating of signal categories.