A Statistical Sparsity-Based Method For Sensor Array Calibration

In Internet-of-Things applications, direction-of-arrival (DOA) techniques play an important role in identifying the direction of the sources. However, the antenna array system is difficult to be precisely calibrated in practical scenarios. In this paper, the DOA estimation problem is considered for the antenna array with unknown errors. In particular, the sparsity in spatial domain is exploited, where the errors are practically assumed to be unknown without any prior information. The proposed algorithm can calibrate the unknown errors in the sensor array and estimate the DOA of the sources simultaneously. Notably, the proposed algorithm is based on a fast sparse Bayesian framework, where the whole process can be carried out with high efficiency. By exploiting sparsity in a statistical manner, the number of sources is not required to be known a priori. The simulated results can validate that the proposed method is capable of obtaining accurate estimation with low computational complexity.