Research On Distributed MIMO Radar Waveform Design And GPU Signal Processing Platform

Distributed MIMO(Multiple Input Multiple Output)radar is an important direction of recent radar research.In the ground-based large-scale distributed MIMO radar system studied in this thesis,each radar node mainly adopts a co-located MIMO working system with a strong anti-intercept ability.This thesis uses genetic algorithms and hybrid coding to generate high-performance orthogonal radar waveforms.Using GPU(Graphic Processing Unit)+ CPU(Central Processing Unit)heterogeneous distributed computing model,using GPU's powerful parallel processing capabilities to accelerate each link of signal processing,allowing multiple GPUs to process multi-node data and speed up signal processing.Aiming at the distributed MIMO radar waveform design and GPU signal processing platform,the main work of this thesis is as follows.(1)In order to make the signals transmitted by the radar nodes orthogonal to each other,two genetic algorithms are used to design the orthogonal binary phase code,which verifies the usability of the designed phase code in terms of auto-correlation and crosscorrelation characteristics;in order to make the distributed radar various waveforms can be adaptively transmitted,the design of mixed coded signals has been studied.For the four signals of binary phase code signal,chirp signal,chirp-binary phase code mixed signal,and binary phase code-chirp mixed signal,from ambiguity function,range resolution,Doppler tolerance,auto-correlation and cross-correlation characteristics,MTD(Moving Target Detection)characteristics are analyzed for the characteristics and applicable scenarios of each signal;then the variable length code transmission method suitable for distributed MIMO radar systems is studied here;combined with the above research,the principle of adaptive waveform selection for distributed MIMO radar system is finally given.(2)Aiming at the problems of a large number of distributed MIMO radar nodes,a large amount of data,and a slow signal processing speed,GPU is used to process node data in parallel and a signal processing platform software architecture design plan is given.A hierarchical,modular,and secure software architecture is designed to ensure the reliability of the GPU signal processing platform.Subsequently,several GPU acceleration methods of single-node beamforming,matched filtering,MTD,and constant false alarm detection are introduced.The acceleration effect of the measured data is analyzed,and the processing speed of the simulation software is compared to verify each module.The acceleration effect.Then it introduces the search synthesis algorithm for multi-node synthesis,analyzes its GPU acceleration method,and detects its acceleration effect.Finally,a terminal display interface based on network data transmission is designed to improve the ease of use of the platform.