Research On Optimization Design And DOA Estimation Method Of Millimeter Wave MIMO Radar Array

Angular resolution is a key performance indicator in MIMO radar systems,and traditional methods increase angular resolution by increasing the number of elements,but it will lead to a significant increase in the integration difficulty and cost of MIMO radar systems.The array sparsity design method can achieve the required beamwidth and angular resolution with fewer array units,thereby saving hardware costs and reducing processing complexity,which is an important way to effectively solve the problem of angle resolution performance and MIMO array cost constraints.The existing MIMO radar array sparse design method is mainly based on intelligent optimization algorithm,and there is a problem of precocious maturity in the process of finding the optimal solution.At the same time,the sparse array causes damage to the matrix structure,which invalidates the existing efficient DOA estimation methods.In view of the above problems,Thesis focuses on the optimization design of MIMO radar sparse array and the fast DOA method based on sparse MIMO array,which is mainly as follows:1.The structure of sparse MIMO array is studied,the signal model of sparse MIMO array is analyzed,and the correspondence between MIMO array and equivalent virtual array is established,which provides a theoretical model basis for subsequent MIMO array optimization.2.An improved adaptive Hamming particle swarm algorithm is proposed,which dynamically adjusts the inertia weight of the particle swarm algorithm by the Hamming distance of the population,solves the precocious problem of the traditional particle swarm algorithm,and models the MIMO array joint optimization problem to make it suitable for the proposed algorithm,so as to solve the MIMO array joint optimization problem by using the Hamming particle swarm algorithm.3.The fast SPICE(Spares iterative covariance estimation)algorithm is proposed,which solves the problem that the sparse array cannot be efficiently estimated due to lack of data through the low-rank completion algorithm.The methods proposed in Thesis have passed the simulation verification,and the optimization effect of the proposed Hamming particle swarm algorithm is improved by5% compared with the common particle swarm algorithm under the same optimization conditions.The fast SPICE algorithm improves the speed by 258% under the premise of ensuring the DOA estimation effect.