A hybrid signal source signal statistics and localisation algorithm

The localisation of mixed signal sources includes the localisation of far-field signal and near-field signal source, which are prone to false alarms (more spectral peaks than signals) or missed alarms (fewer spectral peaks than signals) when using the traditional two-step MUSIC estimation. In order to accurately locate all signal sources’ position, the number of far-field signal and near-field signal sources is first estimated using a signal source number estimation algorithm. Next, a 2D-MUSIC(two-dimensional multiple signal classification) algorithm is used to estimate the positional information and parameters of all near-field signal sources. Based on this, the number of far-field signal sources is obtained by eliminating all near-field signal sources, and the angular parameters of the far-field signal sources are further confirmed by a one-dimensional angular search. Simulation experiments verify the effectiveness of the algorithm. The algorithm compensates for the possible pseudo-spectral peaks of the multiple signal classification algorithm, which can lead to incorrect parameter estimation, by means of a source number estimation algorithm, and laid a foundation for the identification and suppression of ambient signals.