Simple and high-performing DOA estimation by a field of blind beepers

The estimation of the DOA of an acoustic wavefront is considered in the context of the wireless sensor networks (WSNs). An ensemble of simple acoustic antennas, individually with no directional capability, is employed to the estimation aim. The network formed by these antennas (nodes) follows the SENMA paradigm: a mobile rover polls the nodes of the network and implement the estimation task after a certain number of queries of the nodes lying inside its mobile and eccentric field of view. Notwithstanding the DOA-blind character of individual nodes, their aggregation and the directivity of the rover's lobe makes possible inference of the acoustic DOA. The estimation error of the proposed method is close to that of the maximum likelihood (ML) approach, with the remarkable difference that the designed strategy is by far simpler to implement