Bearing Estimation with Randomized Linear Arrays

The source location estimation of multiple signals with a randomized array is implemented in this paper. The incoming signal is sampled non-uniformly in the spatial domain using randomly scattered antenna elements on a linear axis. Mercer's kernel interpolation is performed to map the outputs of the actual array elements in randomized positions to the elements of a virtual uniform array. The sampled signal is passed through the proposed filter and an uniformly sampled equivalent of the received signal is obtained. The interpolated signal is further processed through root-MuSiC to obtain the angles of arrivals of the signals. The mean squared error (MSE) of the proposed estimator is compared to the MSE of the least squares interpolation technique, simulation results are detailed and discussed.