MLANS neural network for track before detect

We have developed a novel approach to enhancing surveillance system capabilities by combining surveillance functions and utilizing all the available information for each function. This approach is based on a previously developed MLANS neural network. The MLANS capability for fuzzy decision logic permits combining such functions as data correlation, detection, and track estimation, or sensor fusion and correlation. This paper considers the problem of concurrently performing detection, correlation, and track initiation for multiple objects in presence of noise or clutter returns. In this case the MLANS estimates track parameters while performing a fuzzy classification of all returns in multiple scans into multiple classes of tracks