Sparse framework for hybrid TDoA/DoA multiple emitter localization

In this paper, we consider the problem of localizing multiple non-collaborative transmitters by a network of distributed sensor nodes. The nodes are equipped with versatile sensing capabilities allowing them to estimate the time differences of arrival (TDoAs) and/or the directions of arrival (DoAs) of the incoming waves. We formulate the localization task as a joint block-sparse recovery problem and develop a framework that allows to accommodate different types of measures, such as the beamformer outputs in case of DoAs or cross-correlation functions in case of TDoAs. We then propose a reduced-size location recovery approach in which we perform multiple location estimations from partial combinations of measures that are later fused together. Our results indicate that in doing so we can achieve estimation performance superior to that of the fully joint recovery, while keeping a lower computational complexity.