Prediction of radar sea clutter based on LSTM

Abstract

Radar sea clutter is the backscattering radar echo of rough sea surface, the research of radar sea clutter is of great significance to national defense construction and national economic development. Sea clutter prediction is also an important point in radar signal processing field. The traditional sea clutter prediction method has lower precision when predicting long-distance sea clutter data, and when the amount of data is large, the time is also lengthened, thereby reducing the efficiency of prediction. In this paper, a new method based on long short-term memory (LSTM) for predicting sea clutter at longer distances in the atmospheric duct environment using near-distance observations is proposed, the principle of LSTM network is introduced, and the factors affecting prediction accuracy are analyzed. The high precision prediction of radar sea clutter by LSTM network is realized. It provide the basis for further work on inversion problem of atmospheric ducts. It also has very important application value for studying the clutter suppression of the radar model and improving the target detection performance of the radar.