The joint calibration method of mutual coupling and gain-and-phase errors based on covariance matching estimation technique (COMET) are proposed for the uniform linear array (ULA).According to the cost function of the COMET and the errors model of the ULA

an alternative iteration algorithm is presented to optimize all the unknown parameters.In order to avoid the complicated step concerning the DOA estimation in each of iteration

an improved alternative iteration algorithm is presented to reduce the computation complexity via making use of the Toeplitz structure of the covariance matrix for the ULA.Compared with the errors calibration methods based on subspace technique

the two novel algorithms are capable of directly exploiting the time-domain statistical characteristic of the sources and are applicable to different Gaussian noise models

e.g.

the case of nonuniform noise power.The simulation experiments validate the effectiveness and advantage of the novel algorithms.