ABSTRACT
Mobile ad hoc networks (MANETs) is a technology that has been developed for real-world applications. The routing information of MANET can be said to be the backbone for the routing process, which also represents the characteristics or behaviors of routing nodes. The performance of MANET can be improved if the routing is done based on nodes’ routing behaviors. Keeping this in view, we have proposed a knowledge model to analyze the routing node behaviors. Our proposed model uses routing attributes to classify the routing nodes based on a fuzzy proximity relation that induces the almost equivalence classes of routing nodes. On imposing an ordering relation on these equivalence classes, we have obtained ordered categorical classes of routing nodes. Further, we use association rules, Bayesian approach, and formal concept analysis (FCA) on ordered categorical classes to determine the behaviors, predict the hidden associations, and predict the implications and dependencies of routing attributes, respectively. Hence, this classification, prediction, and implication of routing nodes’ behavior could lead to proper decision making through which an effective routing model can be developed for MANET.
