Skip to Main Content
Wireless Sensors have been used for environmental monitoring and battlefield surveillance for many years now. Wireless, low-cost, and low-energy sensors are distributed geographically to monitor and report specific effects such as temperature, light, motion and sound. The hardware, network protocols and information retrieval techniques for wireless sensor networks are areas of active research. While outdoor monitoring has been done using wireless sensor networks for some time now, little work has been done in indoor environments. This paper presents a novel approach towards a real time system for indoor surveillance of campus-like environments for loud acoustic events such as gunshots. The work proposes the use of low-cost MICAz motes and a scalable networking approach that ensures low-energy consumption, reliable communication and high data rate. The system uses a hierarchical decision making model to categorize loud acoustic signals, localize the event and trigger an alert system in case of critical events like a gunshot. Several experiments were conducted with real gunshots and other acoustic sources to validate the detection and localization approach.