Minimum-norm with an application to estimation of bearing angles in passive underwater multi-target scenario

摘要

Eigen decomposition methods like minimum-norm, MUSIC are used either for frequency estimation of signals or for finding the bearing angles of the sources. In passive underwater multitarget scenario, the requirement is to find both the targets frequency and location simultaneously. Minimum-norm along with minimum data length (MDL) is used for this purpose. The correlation matrix is constructed and the eigen values and their eigen vectors are found. Then the number of targets are found using MDL criterion. Considering a single location, the array manifold vector is constructed for each frequency. Then the power is estimated using the minimum-norm algorithm. The location (in terms of angle) and frequency of the target are indicated by the maximum power. This procedure is repeated for other locations for detecting other targets.