Efficient Fuzzy Logic based Data Fusion in Wireless Sensor Networks
Arikumar K. Selvaraj, V. Natarajan, L.Sushma Clarence, M. Priyanka · 2016
Development of new type of multi-model sensor technology shows the importance of Wireless Sensor Networks (WSNs) where multi features can be sensed by a single sensor. Wireless sensor networks are mainly characterized by their limited amount of energy and memory. Constructing a network with this kind of sensor will lead to additional challenges with energy and memory constraints of wireless sensor network. The success of the wireless sensor network is mainly depends on maximum lifetime with high quality of services (QoS). Clustering is one of widely used technique to increase the lifetime of the network by utilizing energy in an efficient manner. In clustering each sensor node reported to it is Cluster Head (CH). Data fusion is an important QoS parameter when forming multi-model based wireless sensor networks. In this network, multi-model sensor nodes sensed different features and formed long message that are sent to CH. In this paper, it is proposed a fuzzy based data fusion method named Efficient Fuzzy Logic based Data Fusion in Wireless Sensor Networks (EFLDF). The primary goal of this paper is to increase the QoS in terms of data accuracy, eliminating data redundancy and increase the networks lifetime. This proposed method has three level fuzzy inference System (FIS) to improve QoS. First and second levels of FIS is performed in each sensor node, that able to collect and fuse the data before it is transmitted to the CH. Third level of FIS is done in CH to avoid data redundancy. Simulated results show the proposed method achieved better performances, while comparing with existing data fusion approaches.