Sparse recovery-based DOA estimator with signal-dependent dictionary

A direction-of-arrival (DOA) estimator based on sparse recovery, where the dictionary entries are formed by discretizing the angle space and correspondingly sampling the received signal. Furthermore, to reduce hardware realization complexity, we use the level-triggered sampling (LTS) to sample the signal at different sensors, which typically samples below the Nyquist rate for bursty signals and needs only 1 bit to represent each sample. The sampled signal vectors and dictionary are then obtained through interpolations. Simulation results show that the proposed method, when compared with existing compressive sensing-based methods, achieves better estimation performance with less samples.