Research On High Accuracy And Low Complexity Direction Of Arrival Estimation Algorithm

In Sensor Array Signal Processing,Direction-of-Arrival(DOA)estimation is one of the most important research directions.It is widely used in Military field and national economy such as exploration,radar,navigation,radio astronomy and earthquakes.In the actual DOA estimation system,the related interferometer algorithm is a commonly used direction finding algorithm,which has the advantages of simple algorithm and high sensitivity,and the algorithm has high precision,so it is widely used in multi-channel and single-channel systems.However,in the actual system,the relevant interferometer is affected by many errors,which results in the measured DOA azimuth and the actual wave arrival azimuth.Therefore,it is necessary to perform multi-error modeling analysis and correction for the relevant interferometer algorithm.At present,the direction finding algorithm applied in the actual system is mainly related interferometer algorithm,but the super-resolution direction finding method is a mainstream trend.In the future,the super-resolution direction finding algorithm will be adopted in practical system application,and the random maximum likelihood Algorithm(SML)will be adopted.SML algorithm is an important super-resolution direction finding algorithm.In theory,SML algorithm has the best DOA estimation performance,because the SML algorithm solves the multidimensional nonlinear optimization problem,the computational complexity is very high,it is prevented from being applied in practical systems.Therefore,this paper will analyze and improve the SML algorithm so that the algorithm can be widely applied in practical systems as early as possible.The main research work is as follows:Firstly,according to the various errors generated by the relevant interferometer algorithm in practical system applications,this paper presents a multi-error model for the relevant interferometer algorithm and corrects the error model.The error of the DOA azimuth estimation mainly comes from the error of the frequency measurement of the frequency discriminator,the error of the RF switch during phase switching,the amplitude error obtained during signal measurement,and the proportional relationship between the radius of the element and the signal wavelength.For these four kinds of errors,this paper models and analyzes,and further uses the active correction algorithm to correct.The active correction algorithm used in this paper is based on the active correction of the relevant interferometer algorithm.Since the active correction algorithm is off-line correction,the calculation amount is not considered.Therefore,the MUSIC algorithm is used to solve the matrix.The experimental results show that the azimuth estimation of the signal considering the multi-error model is more accurate than the measurement of the signal azimuth estimation without considering the multi-error model.Aiming at the problem of high computational complexity of SML algorithm,an SML algorithm for initializing space of particle swarm optimization(PSO)algorithm is proposed.The algorithm overcomes a drawback that when the ESPRIT algorithm is used to define the PSO initialization space,the ESPRIT algorithm cannot directly process the signal when the array structure is a non-uniform linear array and the signal is a coherent signal,and a set of preprocessing techniques is required,which increases the complexity of the algorithm calculation.The key point of the proposed algorithm is to use the hypothesis technique to determine the initialization point instead of the ESPRIT algorithm,and to determine the initial solution space of the PSO algorithm by combining the Cramer-Roller Circle(CRB).This method does not need to adopt the pre-processing technique and utilizes the limitation.The algorithm of PSO initialization space greatly reduces the computational complexity of SML algorithm.The experimental results show that the proposed algorithm provides quite good initial values for both coherent and non-coherent conditions.Finally,the algorithm is compared with many existing algorithms,and the effectiveness and accuracy of the proposed algorithm are verified.