Main Article Content
In this research paper, we address the Sybil attack in MANET. Most often the traditional measures against Sybil attack are not applicable in a MANET due to the differences in aims and architectures. The nodes in MANETs are mobile and limited in resources such as battery, bandwidth etc. Due to frequently changing topology it becomes expensive and impractical to authenticate communication and keep track of devices with a centralized server in such network. We proposed a framework which uses nodes attributes in order to judge their behavior. As Sybil nodes consumes data packet by providing wrong routing information repeatedly, we focus on the distinction of the node parameter value of Sybil node and legitimate node. This distinction is made on the basis of nodes attributes and by assigning fuzzy membership values to each node. For this we use fuzzy inference rule and Bezier curve as a tool. Here suspect any node means is to identify those nodes also whose behavior lie in between legitimate and Sybil nodes. In contrast to the traditional crisp logic in which a node can be either Sybil or legitimate fuzzy inference based model focus on the possibility of a node being Sybil. Later we verify their RSP values by drawing a Bezier curve at two different instant of time to identify the Sybil character.