Research On DOA Estimation Method In Radar Communication Integration System

In the vehicle networking environment,using MIMO-OFDM radar communication integration signal,the estimated direction of arrival(DOA)value between the vehicle combined with the distance and speed parameter information can be achieved by vehicle networking location.Thus,the DOA estimation in vehicle networking has important research significance.This paper focuses on the problem of one dimension and two dimension DOA estimation algorithm and its application in vehicle networking environment.According to the requirement of the vehicle networking environment,one dimension and two dimension DOA estimation methods are proposed.The specific research contents are as follows:(1)By building the system modeling of radar communication integration in vehicle networking,which the vehicle networking is at the same plane.The research of one dimensional DOA estimation,according to the requirement of real-time positioning,under receiving small sample size and low signal-to-noise ratio conditions,the performance of the traditional DOA estimation algorithms sharply decline.To sovle these problems,this paper proposes an iterated DOA estimation method which can be adopted in the vehicle networking location in real-time.This method needn't an eigenvalue decomposition or spectral peak searching,and we can update the data covariance matrix to approach to the iterated DOA estimation.(2)Not all of the vehicles are on the same plane,so it is necessary to study the estimation of two dimensional angle.The traditional two dimensional DOA estimation algorithm not only need feature value decomposition and spectral peak searching,but also need additional elevation and azimuth matching.Due to the computational complexity,therefore,traditional two dimensional DOA estimation algorithms are very difficult to be applied in actual vehicle communication system.In view of the actual vehicle networking system,this paper presents a new method of two dimensional DOA estimation which can apply in the vehicle networking,which can estimate elevation and azimuth angles,simultaneously it will realize automatically pair between the elevation and azimuth angle,and the amount of calculation is greatly reduced.Simulation results show,the azimuth angle estimation of vehicle networking in the same plane,the iterated DOA estimation method without characteristic decomposition proposed in chapter three,without spectral peak searching and has more stable performance in small sample number and low signal to noise ratio,computation is much less than the traditional DOA estimation algorithm,which can meet the real-time vehicle networking parameters estimation;The two dimensional DOA estimation method adopted in chapter four can realize automatically pair between the elevation and azimuth angle,and it nnedn't feature decomposition and spectral peak searching,greatly reduces the amount of computation.The simulation results prove the feasibility,robustness and superiority of the method,which lay a theoretical foundation for the real-time location of vehicle networking.