Italian Navy, CSSN ITE “G. Vallauri”, viale Italia 72, Livorno 57127, Italy

In the new world military scenario, potential terrorist action against port structures is of increasing concern. By this way one of the most dangerous situations is represented by individual divers’ attacks which have to be considered as critical as an intrusion conduced by military divers of an enemy nation army. In fact, divers can approach protected littoral areas by exploiting regions inaccessible to acoustic-based systems, such as acoustic shadow zones caused by natural or man made actions (i.e. irregular seabed morphology, obstacles, high reverberation regions, etc…). The magnetic detection method offers a potentially interesting solution to this problem in such regions by integrating acoustic based intruder detection systems. This paper shows the result obtained during the operative test of a port protection underwater system based on a magnetometers’ new self-informed network. The experiment takes place in a geomagnetic space characterized by medium-high environmental noise with a relevant human origin magnetic noise component. The system has two different input signals: the magnetic background field (natural + artificial noise) and a signal composed by the magnetic background field and the magnetic field due to the target (informative signal). The system uses the first signal as filter for the second one to detect the target magnetic signal. The effectiveness of this procedure is related to the position of the magnetic field observation points (reference devices and sentinel devices). The sentinel devices must obtain correlation in the noise observations and de-correlation in the target signal observations. The system, during four tries of intrusion, has correctly detected all magnetic signals generated by divers. The processing of magnetic signals has been enhanced by implementing an amplitude threshold to cut the passive energy components, which are those components that don’t contain useful information about diver passages. The use of this Passive Energy Cutter (PEC) increases the reliability of the system in terms of detection probability versus false alarm probability, as evidenced by the comparison of the ROC curves for the system with and without the PEC use.