Automated Processing For Distributed Undersea Sensor Systems

Future expeditionary forces in littoral environments will confront the submarine threat by using networks of small, easily deployable arrays. For small arrays passive detection ranges can be too short to support practical system concepts. Limited detection ranges are partially due to source-motion-limited integration times. Detection of quiet submarines requires longer integration times than are allowed by quasi-static assumptions of beamform-detect-track methods. George Mason University(GMU) has developed a broad-band passive tracking algorithm that automatically associates source spectra with tracks. The method can exploit very long integration times. RobTre proposes to develop an algorithm for robust submarine track detection which exploits the GMU tracking algorithms extended integration and signal separation capabilities. In Phase I, simulations will demonstrate robust track detection in the presence of a loud moving surface ship. During Phase II, the method will be extended to incorporate advanced submarine broad-band signature characteristics for improved noise rejection. Additionally, a robust shallow water signal model will be incorporated that increases signal gain and enables motion detection through multipath differential Doppler without the need for matched field processing. The effectiveness of the new method will also be demonstrated using data in Phase II.