Antenna array self-calibration algorithm with sensor location errors

The antenna array characteristics like sensor location, gain and phase response are rarely perfectly known in realistic situations. Self-calibration algorithms estimate both source DOAs and perturbed array response vector parameters simultaneously. In this paper calibration errors are usually assumed Io be small and a first order approximation to the perturbed array response vector is often used to simplify the estimation procedure. We introduce two classes of self-calibration algorithms for solving the sensor location error problem. This paper investigates the performance of these algorithms. Simulation results show that iterative approach and MAP-NSF algorithm are much better than MUSIC algorithm, and the iterative approach is preferred. But the iterative approach is very sensitive to initialization. A novel algorithm, using the MAP-NSF algorithm results as the input to the iterative approach, is proposed to eliminate sensor location errors and remove the sensitivity to initialization. Further simulation results demonstrate the improved performance of the new approach.