Analysis of the performance and sensitivity of an eigenspace-based interference canceler

Eigenspace-based interference cancelers (EICs) possess the advantages of providing maximal suppression of interference with fast convergence over conventional adaptive beamformers. However, the performance and sensitivity to steering angle error of EICs have not been analyzed due to the use of a signal blocking matrix. We first present a technique to construct a positive definite matrix based on the signal blocking matrix and then use this matrix to compensate the effect of the signal blocking matrix on the sensor noise received by an EIC. Therefore, the interference subspace required for finding the optimal weight vector can be obtained using conventional eigenvalue decomposition (EVD). Moreover, the performance and sensitivity to the steering angle error of the EIC can be analyzed. Simulation examples are provided for confirming the theoretical results.