Theoretical and experimental studies on high resolution iMP algorithm for bearing estimation

This paper involves theoretical and experimental studies on the high resolution IMP algorithm. Two modification methods are proposed for the improvement of convergence speed of IMP, one is to reduce the number of iterations required in the original IMP algorithm and the other is to speed up each iteration involved. Applying these two methods to the original IMP algorithm increases the convergence speed by 50%, verified by computer simulations. In-door experiments are conducted in a water tank for studying the performance of IMP in practical situations. By calibrating with measured array manifold of the experimental array, the IMP algorithm separates two strongly correlated sources with different strengths reliably and estimates their directions correctly.