Forest Fire Detection in Wireless Sensor Network Using Fuzzy Logic

Abstract

The multi-purpose integrated homeland surveillance security systems are usually located in remote areas. Intelligent decision making (IDM) capability emerges as the primary feature in the realization of the T/R architecture. The aim of employing IDM is two-fold. First is to save energy, as the system operation is desired to be autonomous based on the available solar power and the corresponding battery-bank. Second is to activate the necessary action(s) required based on the pre-defined sensitivity levels. The current work is focused on the second aim using the pre-defined sensitivity levels. We propose to use a wireless sensor network (WSN) for data harvesting to be used as raw input data into our control system. Fire detection has been chosen to illustrate the IDM capability of the system. A Fuzzy Logic algorithm is developed using five membership functions as temperature, smoke, light, humidity and distance. Simulation results for the probability of fire based on the fuzzy rules using the status of the membership functions are presented in the paper.