Radar Data Processing Based On TMS320C6678

Radar data processing technology is an important part of radar system.Since the real target is usually interfered by clutter,weather and noise,it needs advanced radar data processing technology to track,filter,interconnect,smooth and predict the target data.This technology can achieve the effective elimination of clutter and noise suppression effect and real-time tracking of the target and form a stable target track.This paper combines "XXX array radar" project to design and realize a multi-core DSP radar data processing system based on TMS320C6678.In this paper,Kalman filtering algorithm,track association,track extrapolation,track start and track die in radar data processing are analyzed firstly,in which the track correlation introduces radial velocity observation information,and the track starts with the method of multiple hypothesis track initiation was adopted to effectively improve the tracking performance of the system.Combined with the phased array radar data processing project,the simulation results of the entire process in the Matlab environment are given.Then the overall framework of TMS320C6678 hardware platform is studied,including SRIO,Ethernet,GPIO and EDMA3 controller.The software and debug of the interface and EDMA3 controller are also verified.The reliability and correctness of the interface transmission data and the matrix transposition of EDMA3 controller are verified.In this paper,the project based on two-core radar data processing algorithm is proposed to achieve the program,the inter-core communication,inter-core synchronization and Cache consistency issues are studied.In this scheme,the main coreO is responsible for receiving data and finishing sending data.Corel mainly implements the functions of preprocessing,tracking,filtering,interconnection,smoothing,and prediction.The algorithm time and resource analysis results show that the algorithm meets the real-time requirements of the system.Finally,based on TMS320C6678,the algorithm is debugged,and the radar data processing module and MATLAB results are compared to ensure the correctness of the algorithm.