Direct Position Determination of Quasi-Stationary Sources Based on Virtual Array Synthesis of Distributed Nested Arrays

For the direct position determination (DPD) problem of quasi-stationary sources (QSS), a new DPD algorithm based on virtual array synthesis of distributed nested arrays is proposed. Firstly, multiple nested arrays are virtualized and synthesized into a large aperture virtual array based on only the signal powers. This virtual array model eliminates the different delays incurred by different receiving arrays, thereby avoiding synchronization errors resulting from unsynchronized clocks among the arrays. Subsequently, leveraging the generalized stationarity of QSS, the received signals corresponding to the synthetic virtual array from different time frames are superimposed together to form the final received data model. Finally, dimension reduction is performed by utilizing the Lagrange multiplier method on the cost function constructed by subspace methods, avoiding multidimensional searches for spatial coordinates and signal amplitude attenuation coefficients, which greatly reduces computational complexity. Through simulation comparisons, the algorithm exhibits significant advantages in terms of localization accuracy and the maximum number of identifiable sources.