Research Of DOA Estimation Algorithms Based On Cross Correlation MUSIC

Direction of Arrival(DOA) estimation is an important research subject of array signal processing. The basic purpose of DOA estimation is to confirm the spatial positions of signals which we are interested in by a group of antennas with specific construction. The DOA estimation algorithm is becoming more and more complete through decades of research, and the DOA estimation algorithms based on subspace is the most. The DOA estimation algorithms based on subspace consisted of two kind of algorithms, which are the MUSIC algorithms and the ESPRIT algorithms. The MUSIC algorithms divide the receiving data into two orthogonal subspace, the signals subspace and the noise subspace, and use its characteristic to assess parameters. The ESPRIT algorithms use the characteristic of rotation invariant subspace to assess parameters. This paper mainly analyzes these two kind of algorithms, and use MATLAB software to do simulation experiment, puts forward two new points based on the previous theory. The main work and innovations are as follows:1. Analyze four typical antennas array: Uniform Liner Array, Uniform Circular Array, Uniform Double Parallel Array, Uniform L Type Array. Deduced the array flow matrix of these for antennas array when there are multiple signals.2. Deduced and prove the CBF algorithm, the MVDR algorithm, the MUSIC algorithm, the ESPRIT algorithm, and use MATLAB software to analyze the DOA estimation performance for uncorrelated signals. Put forward the CBF-MUSIC algorithm, and confirm that it has good performance for uncorrelated signals and general performance for correlated signals.3. Analyze two common decorrelated algorithms, the FBSS algorithms and the ESVD algorithm. Put forward the MM-MUSIC algorithm based on cross correlation MUSIC algorithm, and prove it is a bit better through simulation experiment.4. Analyze 2D DOA estimation problem, analyze 2D-ESPRIT algorithms based on Uniform Double Parallel Array and 2D-MUSIC algorithms based on Uniform L Type Array. Introduce one new antennas array, which can transform 2D DOA estimation to 1D DOA estimation.