Research On Array Signal Multi-parameter Joint Estimation Algorithm

Electromagnetic vector sensor is a new type of polarization sensitive array consisting of six array elements, it is widely used in the systems whose transmitting and receiving carrier are the electromagnetic wave. Compared with traditional scalar sensor, electromagnetic vector sensor can receive all the electromagnetic field components, it also has the advantages that there exists the orthogonality between the electromagnetic field components. However, the existing signal processing techniques in view of the electromagnetic vector sensor are mostly built by the original of scalar sensor signal processing technology, which could not fully reflect the advantage of the electromagnetic vector sensor. Therefore, on the basis of the research achievements of many scholars, this paper introduces some cutting-edge theories to improve the signal processing algorithms of electromagnetic vector sensor, the proposed algorithms make good use of the advantages of the electromagnetic vector sensor. The key innovations in this thesis are as follows:This paper firstly introduces the basic theory of the quaternion and quaternion matrix, and then establishes the electromagnetic vector sensor array output signal model of the quaternion. Using the expression of quaternion model, the proposed quaternionESPRIT can estimate the DOA and polarization parameters. Because the quaternion model can make full use of the local vector characteristics of electromagnetic vector sensor and the three-dimensional structure characteristic of the whole array, the quaternion method has the adva ntages of smaller computational complexity and higher estimation precision than that of the long-vector method. The simulations of the uniform linear array and uniform circular array are also conducted. The simulation results have verified the correctness and effectiveness of the proposed method.This paper links the multi-dimensional matrix decomposition to the basic theory of the PARAFAC model, and establishes a three-dimensional matrix model of the electromagnetic vector sensor array. This link facilita tes the three dimensional matrix decomposition and identifying problems by using the PARAFAC model, and then estimates the DO A and polarization parameter at the same time. The three-dimensional matrix model can fully dig the array data high dimensional information. Thedecomposition result of PARAFAC model is unique without considering the condition of column permutation ambiguity and scale fuzzy. Therefore, the PARAFAC method has the higher estimation precision than that of the ESPRIT method. The simulatio ns of the uniform planar array and uniform cylindrical array are also made. The simulation results have verified the correctness and effectiveness of the proposed method.This paper introduces the basic principle of compressed sensing theory. When the DOA parameters are estimated using the compressed sensing theory, there exists the issue of parameter mismatch. To solve the above issue, a new dimension reduction compressed sensing parameters matching algorithm is proposed. The proposed algorithm provides the good idea that the two-dimensional compressed sensing is reduced to two one-dimensional compressed sensing, thus the computational complexity is greatly decreased. Finally the two-dimensional DO A estimation is obtained by matching operation, which applies the coherent or incoherent signal. The proposed algorithm also has the advantage of high precision estimation perfo rmance. The simulations of the L-shaped array configuration are also made. The simulation result has verified the correctness and validity of the algorithm.