Design and implementation of flood detector using wireless sensor network with mamdani's fuzzy logic method

Abstract

In this paper, we build the prototype system using Wireless Sensor Network (WSN) for flood detector for flooding area. In a flooding area, power consumption and reliability of network become an important thing. Therefore, the performance of devices and algorithm of this system must work properly. The system is created using Fuzzy Logic to calculate the output, in order to make the system work properly. With Zigbee technology, WSN can be useful for monitoring a small area with many nodes. WSN can be combined with another application, e.g. cloud, android, virtual private network, etc. The objective of this paper is to design flood detector system with WSN on two scenarios that are single hop and multi hop, then find two of these which the best. Single hop means no router node between coordinator node and end node and multi hop means there is router node between coordinator node and end node. After some test, this system has a maximum range in single hop at a distance of 95.1 meters and 185.5 meters when using multi hop's scenario. Error for fuzzy logic compare between manual calculation and prototype system is 3.04417493% and error overall system for its result is 5%. In the examination of the quality of the network, values of throughput obtained fairly stable but the result of delay values is unstable, the further the distance the higher the values of delay, on the contrary values of throughput getting lower. The power consumption is only 0.083 Watt/hour when the system in single hop and 0.06525 Watt/hour when the system in multi hop.