A Radar Signal Sorting Method Based On U-Net Image Semantic Segmentation Technology

Radar signal sorting plays an important role in the information age,especially in electronic warfare.The rapid development of deep learning technology has brought the possibility of innovation to the research of radar signal sorting.In order to observe the position characteristics of radar signals and sort them better,this paper proposes a method of radar signal sorting based on U-Net image semantic segmentation.The core idea of this method is to draw the mixed radar signals into images,and then use the U-Net network model to carry out the semantic segmentation of images,so as to complete the task of radar signal sorting.First of all,from the practical significance of radar signal sorting,this paper introduces the role of radar signal sorting in military application and traditional radar signal sorting algorithm,and discusses the current research status of radar signal sorting at home and abroad.The traditional radar signal sorting algorithm can achieve the purpose of sorting only for a certain type of radar signal,or it can only sort radar signal in a specific environment,which has many limitations.Secondly,it introduces two kinds of methods of image semantic segmentation,which are traditional image segmentation method and image segmentation method based on depth learning.It focuses on several depth learning image segmentation methods and summarizes their advantages and disadvantages.Then,this paper proposes a method of radar signal sorting based on U-Net image semantic segmentation technology,according to the pulse repetition interval(PRI)frequency matrix of radar signal draws the radar signal image,draws the annotation image according to the annotation matrix,improves the traditional U-Net network,trains the radar signal images and the annotation images with the improved U-Net network model,tests the model with different radar signal images,and generates theradar signal segmentation result maps.Finally,this paper uses three different groups of radar signal simulation datas to carry out experiments,analyzes the experimental results,verifies the effectiveness of the proposed radar signal sorting method in signal sorting,summarizes the shortcomings of the current model and future research work,and makes a prospect for it.