Multi-Parameter Joint Estimation Algorithm Of Under-Sampled Array Signal

With the continuous development of wireless communication,the abundance of access devices and the large broadband required for information transmission have made the problem of spectrum congestion increasingly severe.Under the framework of cognitive radio that can dynamically allocate spectrum resources,the frequency and DOA joint spectrum estimation based on array technology can perceive spectrum holes from the two dimensions of frequency domain and space domain,so as to obtain more access opportunities for users and effectively alleviate the spectrum crowding problem.On the other hand,spectral sensing technology based on under-sampling can break through the Nyquist sampling theorem and solve the problem of wideband signal sampling.Therefore,this paper systematically studies the DOA estimation problem under under-sampling and the joint estimation of multi-band signals,aiming to solve the problems of accuracy,robustness,and scene complexity of the multi-parameter joint estimation of under-sampled array signals.The main contributions of this paper are as follows:First,this paper studies the DOA estimation problem under sparse arrays.In view of the high computational complexity of existing algorithms as the number of antennas and snapshots increase,this paper uses one-bit quantization technology and the framework of compressed sensing to transform the DOA estimation problem into a classification problem,and proposes a DOA estimation algorithm based on classification under sparse arrays,and the Cramer-Rao bound of the model is derived.Under this framework,classic binary classification algorithms in machine learning,such as logistic regression,support vector machines,etc.,can be used to solve DOA estimation problem.What's more,the estimated angle can be further corrected through the idea of grid refinement.Simulation results show that the proposed algorithm is superior to existing algorithms in terms of accuracy and resolution.Next,in order to solve the problem of joint spectrum estimation of frequency and DOA of the under-sampled far-field multi-band signals,this paper combines the generalized L-shaped array with the Modulatd Wideband Converter(MWC).When the L-shaped array in the system structure is uniformly arranged,this article introduces two classic joint spectrum estimation algorithms,including Joint ESPRIT algorithm and CS algorithm.Later,in order to solve the problem of high hardware complexity of existing uniform array algorithms,by introducing spatial under-sampling technology,that is,sparse array,this paper proposes a joint estimation algorithm of frequency and DOA under spatio-temporal under-sampling.The simulation results show that the proposed algorithm forms an equivalent virtual array with increased array aperture,thus,the performance of joint estimation of frequency and DOA is better than ESPRIT algorithm and CS algorithm based on uniform array.Finally,based on the research foundation of the first two points,this paper proposes a three-dimensional parameter and four-dimensional parameter joint estimation framework for under-sampled mixed far-field and near-field multi-band signals,and derives the performance bounds of the proposed algorithm.For different estimation targets and scenarios,this paper designs corresponding undersampling structures.The system structures used in the joint estimation of the threedimensional and four-dimensional parameters of the mixed field are the symmetric ULA-MWC and the symmetric cross array-MWC system structure respectively.Under this framework,this paper uses the covariance algorithm,PM algorithm,and JAFE algorithm to obtain the estimation of the carrier frequency and its mixed signal subspace first,and then uses the estimated mixed signal subspace combined with subspace theory,compressed sensing theory,and oblique projection technology to estimate the corresponding far-field and near-field parameters.Compared with the existing algorithm,the proposed algorithm framework has only one unknown parameter in each link,and the carrier frequency and mixed signal subspace are automatically paired,which solves the problem of the traditional multi-parameter joint estimation algorithm due to multi-dimensional searching,parameter pairing,and the problem of high computational complexity caused by matrix eigenvalue decomposition.The simulation results show that with the increase of system structure delay channels,the performance of the algorithm in joint parameter estimation is getting better,that is,the JAFE algorithm is better than the PM algorithm,and the PM algorithm is better than the Covariance algorithm.