چكيده به لاتين
One of the important topics in the field of radar MIMO, estimating the angle of the target. Various methods for improving the estimation of MIMO radar angle is used. One of these methods, the use of compressive sensing theory. This sampling at a rate much lower than the Nyquist rate possible and can be measured with the help of a few examples, the signal is sparse or thinning almost be restored. In this thesis, using compressive sensing Bayesian hierarchical clustering based on MIMO radar to estimate to improve our performance. In the proposed method based on sparse signal coefficients Myarhdaql the distance between the data aggregated hierarchical model is created. The number of non-zero coefficients, depending on the type of clustering is determined by the use of soft zero. Bernoulli random vector z can be used to show non-zero elements. The distribution function to be determined by the type of clustering. The combination of sparse signal distribution function coefficients and Bernoulli random vector, probability distribution function of the previous spike-and-slab form. To estimate the sparse signal and angle of the targets of the previous spike-and-slab probability distribution function we use. The proposed method will be compared with Bayesian compressive Sensing method and angle of targets using a sparse signal reconstruction of the two methods is estimated. Standardized methods to compare the mean square error of the angle. The proposed method estimates the target angle, relative to the size of the Bayesian Compressed Sensing method improves 0.35. In terms of runtime simulation, the proposed method to size 0.22 Minutes later will result. The proposed method can separate targets with different angle of 0.3 degrees and estimate its direction.
