Wireless Sensor Networks and Audio Signal Processing for Homeland Security

A plethora of solutions to homeland security problems have been proposed, during the last years, by academia, national governments, and industries. In this chapter, we focus on homeland security solutions based on efficient Wireless Sensor Network (WSN)-based audio signal pattern recognition. This is of interest for efficient surveillance of the perimeters of large areas, in order to detect the intrusion of humans or vehicles. We first propose a simple time domain approach to signal pattern detection and its commercial application through Unattended Ground Sensors (UGSs). Then, we extend this approach in the direction of a hybrid time-frequency approach, obtaining a very good performance yet with limited complexity