Research On Target Parameter Decoupling And Estimation Technology Based On Frequency Diverse Array

Frequency Diverse Array(FDA),which was put forward in recent years,has been highly concerned by researchers at home and abroad.Compared with the traditional phased array(PA),FDA applies an additional frequency offset to each element of FDA.In other words,the transmission carrier frequency of FDA is not equal,and the transmission signal of phased array has the same carrier frequency.In traditional sense,the linear frequency offset of FDA will lead to the direction pattern of FDA has the direction angle distance dependence.However,FDA pattern is coupled in the dimension of distance and angle,and the beam energy is a continuous point.This coupling phenomenon is the difficulty of the target parameter estimation of FDA.Therefore,this paper mainly studies the decoupled method of the target parameters of FDA,aiming at the distance angle coupling problem of FDA pattern.At the same time,the algorithm of target parameter estimation is optimized.Simulation results show that this algorithm has higher estimation accuracy than traditional MUSIC algorithm.The main research contents are as follows:Firstly,the traditional linear array is used to study the azimuth distance coupling problem of FDA,and its pattern is deduced.For the two-dimensional coupling problem,the basic idea of decoupling of nonlinear array or nonlinear frequency offset is proposed.The advantages and disadvantages of MUSIC,TLS-ESPRIT and Root-MUSIC algorithms in the estimation of FDA target parameters are analyzed,and a new reduced dimension MUSIC algorithm is proposed.Secondly,a specific decoupling scheme is proposed for the azimuthal distance coupling problem of FDA.Firstly,by changing the array structure to decouple,the L-type array structure with z-axis as phased array and x-axis as FDA is proposed,and compared with multi carrier array and differential array.Simulation analysis shows that the L-type array structure not only has the best estimation performance,but also has no complexity and limitation in practical application.Finally,the nonlinear frequency offset decoupling scheme is introduced,and the estimation performance of the frequency offset in the form of power function,log function and Costas sequence is analyzed by simulation.Thirdly,aiming at the problem of poor estimation performance in the case of short snapshot or even single snapshot,this paper combines the idea of sparse reconstruction of compressed perception to accurately estimate the signal azimuth information with less sampling data.Two reconstruction algorithms,orthogonal matching pursuit(OMP)and subspace pursuit(SP),are simulated and analyzed.According to the mismatching of grid,an off grid algorithm based on Sparse Bayes(OGSBI)is proposed.The simulation results show that both the OMP and SP algorithm are better than the traditional estimation algorithm in the case of single snapshot,but the error is still a little larger,and the OGSBI algorithm can well estimate the target when the grid is mismatched.At the same time,the l₁-SVD algorithm based on preprocessing is proposed to further optimize the performance of single snapshot estimation.Simulation results show that the estimation error of this algorithm is better than other reconstruction algorithms used in this paper.Finally,aiming at the problem of poor performance of angle-range estimation in the case of low SNR,this paper applies higher-order cumulant to the target parameter estimation of FDA radar.In addition,MUSIC-Like algorithm is proposed,which makes use of the fact that the fourth-order cumulant is not sensitive to Gaussian signal,so as to achieve the purpose of noise suppression.At the same time,higher-order cumulant can virtual more array elements.It has the advantage of expanding the array to make up for the poor estimation performance of the existing algorithm at low SNR.The simulation results show that the higher-order cumulant is very effective in solving the problem of low SNR and has good estimation performance.