Anti-interference Simulation Platform Construction And Key Technology Research

The current air defense system is facing an increasingly complex combat environment,and the development of Anti-jamming technology is urgent.Therefore,designing a comprehensive anti-jamming simulation platform with scalability and close to the real combat scenario has strong scientific research and engineering application value.This paper designs a distributed collaborative anti-jamming simulation platform based on RTX and reflective memory technology.We complete the overall architecture design of the distributed collaborative anti-jamming simulation platform.And realize the platform framework using c ++ language.From the simulation results and analysis,we find: the system latency of the simulation platform is less than 10 us,and the data transmission speed exceeds800 Mbps.The platform has excellent real-time performance.Then we focus on the simulation of radar signal processing algorithm,and use multi-thread processing method to achieve signal processing acceleration The signal processing time is shortened to one tenth of the original average,making the simulation closer to the real scene.Validated on the designed anti-jamming simulation platform,we find the modules of the simulation platform can work together to track the target stably and accuratelyFinally,the intensive pattern recognition method based on deep learning is researched deeply.After sampling and normalizing the MTD output image of the PD signal of the target and 9 single-mode interferences,we use the designed four-layer convolutional neural network to train and recognize thiese signals.At first,we use global image recognition and detail image recognition methods to train the signals respectively.And based on this,we propose a secondary detection method that using global image recognition to isolate most of the interference,and using detailed image recognition to isolate the interference that is difficult to identify.The results show that after using the secondary detection method,the recognition accuracy of all interferences is more than 96%.