Research And Application Of DOA Estimation Method Based On Subspace

As one of the key issues of array signal processing, Direction of Arrival(DOA) estimation has found wide applications in many fields, such as radar, communications, sonar, seismology, radio astronomy and biomedicine engineering, etc. The algorithms based on subspace play an important role in DOA estimation for their excellent performance and high resolution. Multiple Signal Classification(MUSIC) and Estimate Signal Parameter Via Rotational Invariance(ESPRIT) are two typical algorithms of DOA estimation. This paper designs a Matlab simulation on the analysis of two algorithms and analyzes their performance.The paper presents the advantages and prospect of the polarization sensitive array and shows how subspace methods be used to estimate the DOA parameters with a polarization sensitive array. The performance of polarization sensitive array is better than linear array but the estimation of DOA parameters is more complicated because more dimension and polarization parameters are imported. In this paper, the model of polarization sensitive array signal is established according to the different distribution of the array.The three-dimensional modulation signal is modulated by the signal amplitude, polarization angle and polarization phase angle of an electromagnetic signal. The performance of the communication system is effectively improved by extending the dimension of modulation parameters and full use of the polarization information of electromagnetic waves. Based on the character of three-dimensional modulated signal, a novel approach to the demodulation of three-dimensional modulated signal by MUSIC algorithm is presented in this paper. By searching the space spectrum of the received signal, the polarization convergence angle and polarization phase angle were estimated, thus, the baseband signal was recovered. Both the analysis and the simulation results show that the proposed method can effectively receive the three-dimensional Polarization Quadrature Amplitude Modulation(PQAM) signal. This method also has high resolution and accuracy, and gets lower Bit Error Rate and estimation error in the conditions of low SNR.According to the defect of large computation and complexity of traditional algorithm, a novel approach to the fast demodulation of three-dimensional modulated signal was presented. The method was based on the MUSIC algorithm. Firstly, the covariance matrix of the array data was set up by using a single snapshot and secondly the signal space spectrum was decided according to the prior information of signal, then the polarization convergence angle and polarization phase angle were estimated, thus, the baseband signal was recovered. Simulation results indicate that the proposed method avoids spectral search, further reduces the computational complexity and suits for the modulation of fast moving signals.