Signal direction of arrival estimation based high-order cumulants and a partially calibrated array

For the problem of obtaining high-resolution direction of arrival (DOA) in the case of the angularly dependent array uncertainties and non-Gaussian sources, in this paper, a new method of signal direction of arrival estimation based on high-order cumulants and a partially calibrated array (CPCA) is proposed. In this algorithm, the high-order cumulants of received signal is accounted, and the corresponding signal model is derived by characteristic of the gain-phase errors. Then, using the idea of ISM, the signal model is transformed in consideration of the number of uncalibrated and well-Calibrated elements, and the matrix transform is also obtained by all of gain-phase parameters which is product by Kronecker operation. Finally, the noise subspace is adjusted according to the order of elements of steering vector. This method adapts to arbitrary array, needs relatively small computational capacity and one dimension searching. The computer simulation results illustrated that the new method can eliminate gain-phase errors and color noise, and the system stability.