Covariance matrix based fast smoothed sparse DOA estimation with partly calibrated array

Direction estimation using a two-level nested array with gain–phase error calibration

“…The fast-smoothed sparse method works well with partially calibrated arrays, however, all other methods need no calibration sources. The resolution performance in [6] is better compared to [7] and [8], and the method proposed also ensures good resolution performance. In the proposed method, gain and phase errors are considered together as the error matrix which is minimized whereas methods in [7] and [8] calculate the gain and phase errors separately and calibrated.…”

“…The eigendecomposition of R zerr given in (6). The large eigenvalues corresponding to signal and small eigenvalues give information of noise.…”

“…The algorithm for the proposed method is given in Table 1. The proposed calibration method is compared with covariance matrix-based fast-smoothed sparse method, eigenstructure method, and Hadamard product method proposed in [6][7][8], respectively. The proposed method is to be applied to a nested array for gain and phase calibration earlier.…”

“…Calibration using partially calibrated sources and with a single high-power calibration source is discussed in [5]. A spatial-smoothing-based sparse reconstruction method is used for a partly calibrated array that gives better performance [6]. DOA estimation of unknown directions of sources in the presence of model errors is challenging in such scenarios, self-calibration algorithms will be useful.…”

“…The technique suggested here is based on the eigenstructure of the complex array output covariance matrix. The proposed method performance is compared with the existing fast-smoothed sparse method [6] and eigenstructure-based methods [7,8]. The minimization of gain and phase errors discussed in [9] for a uniform circular array is the inspiration for the application of the proposed method in nested arrays.…”