Skip to Main Content
A computationally efficient algorithm for multiple source localization, using the expectation-maximization (EM) algorithm, for the wideband sources in the near field of a sensor array/area, is presented. The basic idea is to decompose the observed sensor data, which is a superimposition of multiple sources, into individual components in the frequency domain and then estimate the corresponding location parameters associated with each component separately. Instead of the conventional alternating projection method, we propose to adopt the EM algorithm in this paper; our method involves two steps, namely Expectation (E-step) and Maximization (M-step). In the E-step, the individual incident source waveforms are estimated. Then, in the M-step, the maximum likelihood estimates of the source location parameters are obtained. These two steps are executed iteratively and alternatively until the pre-defined convergence is reached. The computational complexity comparison between our proposed EM algorithm and the existing alternating projection scheme is investigated. It is shown through Monte Carlo simulations that the computational complexity of the proposed EM algorithm is significantly lower than that of the conventional alternating projection algorithm.