Abstract-Directed Diffusion is a Data Centric routing protocol. In data centric protocols data is named based on the attribute-value pairs. Directed diffusion protocol includes flooding the interested data, establishing direction paths by using gradients set-up and selecting one or more direction paths to reinforce the data stream. The main difficulty of Directed Diffusion is flooding in which there is some communication overhead with some nodes which are naturally unable to coordinate in monitoring the interested event. We use a fuzzy logic controller to reduce the communication overheads during flooding and routing the data stream from the source(s) to a sink node. The fuzzy logic controller evaluates the potentiality of intermediate nodes to coordinate in mission, based on each node's traffic load, energy residual and size of data which can be maintained. The simulation results show that new Fuzzy diffusion approach produces at most 56.5% and in average 28.81% successful results and has improved the total energy residual at least 19.38% and in average 1888.87% in comparison with directed diffusion protocol.Index Terms-Wireless sensor networks, fuzzy logic, directed diffusion, energy efficiency.
I. INTRODUCTIONA Wireless sensor networks (WSN) contain multitude tiny nodes in which various sensors, actuators, processing and communication units exist. These nodes interact highly with physical environment providing continual unattended monitoring of natural or man-made phenomena and less accessible regions such as poisonous or chemical areas in industries, aircraft interiors and early warning of natural disasters [1]- [2]. These numerous sensing nodes distributed in an area can coordinate with each other and constitute a distributed network to achieve a larger sensing task [3].Wireless sensor nodes are highly restricted by energy level of nodes. Although the recent advances in MEMS and VLSI technologies can lead to building very small and energy efficient sensor nodes enabling to do processing and computation, the demand on decreasing the energy consumption in nodes is still crucial even in some situations where the renewable energy sources (e.g. solar panel) are provided.Besides hardware optimizations, energy efficient routing methods act a significant role in improving the routing energy efficiency.From a point of view, routing methods are categorized into Data Centric (DC) and Address Centric (DC) [4]. DC protocols are significantly different from AC protocols where each node has a unique ID, the routing is end-to-end and each source independently send data along the shortest path to sink based on the routes the queries took [4]. However, in DC protocols there is no globally unique ID. The nodes are capacitated to handle the content of the data and perform some data aggregation on the sensed data which is passed via neighbor nodes.Data aggregation (the combination of data from different sources [4]) dedicates many advantages to Data Centric Models. Of the advantages is eliminating the redundancy in intermediate...