Automatic identification of specular detections in multistatic sonar systems

Multistatic systems have the potential to provide increased ASW performance. However, effective fusion and tracking of multistatic active sonar contacts is challenging, due to high levels of false alarm clutter present on all sonar nodes. Such false alarms often overload the sensor-to-fusion-center communications links and fusion/tracking processes, thereby producing too many false tracks. A system concept referred to as Specular-Cued Surveillance Web (SPECSweb) has the potential to solve this overloading problem. This is done through the exploitation of very strong specular echoes from targets operating within a geometrically diverse sensor field. The strong specular echoes are used as cues for track initiation and track holding through selective extraction of additional detections stored locally on the individual sonar nodes. This approach significantly reduces the data rate at the input to the fusion/tracking algorithm, and reduces node-to-fusion-center communication link throughput requirements. A SPECSweb information fusion and target tracking algorithm has been designed, which has shown to be effectively in providing this unloading. Previously, the identification of specular detections was made by simply setting a higher-than-normal SNR detection threshold at the tracker input. This paper presents an alternate approach for detecting specular echoes, which is based on the identification of amplitude changes in the detection levels corresponding to a target passing through the specular geometric condition. The approach is data-adaptive and more robust that the method used previously. The algorithm is described and factors important for parameter value selection are discussed. The proposed identification scheme is applied to multistatic data, showing its effectiveness.