Utilization of wireless sensor network technology can improve the anticipation of the occurrence of fire hazards in the smart home; this is done by replacing the human task in monitoring the situation around the home by using multiple sensors which can directly interact with the environment. The goal of this paper is building the early fire detection systems on smart home-based wireless sensor network monitoring. Sensors are used to detect the level of fire danger include temperature sensors, humidity, carbon monoxide, and smoke. The system use fuzzy logic inference systems to process the data from the four sensors in order to improve the reliability and accuracy of the information provided for the system that will give warning to users. Based on the results obtained, the system has been able to give an alarm warning to users appropriately. The proposed system also implement sleep scheduling method in the system which can improve the delivery efficiency of data packets and can reduce battery resource usage.