Quality assessment for signal detection in underwater acoustics applications

The paper presents an optimization framework for quality assessment applied to a communication system. It is described as a network of processing modules, each of which characterized by a set of parameters influencing the quality of local processing and, as a consequence, the quality of the whole system results. In particular, the detection task is considered: the most significant stets of parameters influencing quality inside each of the system modules are selected and a quality-assessment criterion is developed in order to determine which values of these parameters allow one to reach optimized signal detection. A set of detection methods is examined and compared in order to establish their feasibility and performances by considering detection probability values as the selected parameters vary. The developed quality assessment framework was applied to a real case of detection of underwater acoustic signals in non Gaussian noise.