A New Self-Organizing Mobile Ad Hoc Network Using Case Base Reasoning

A New Self-Organizing Mobile Ad Hoc Network Using Case Base Reasoning

Abstract

This paper proposes an intelligent weight based clustering algorithm for mobile ad hoc networks. The proposed algorithm takes into account transmission power, battery power, Tnb (connection duration with neighbors), Dnb (Average of distance with neighbors in Tnb), and the degree of a node for forming clusters. New Parameters of Tnb and Dnb are adopted as two metric to elect ClusterHead (CH) in this paper. Using the first parameter increases the stability of cluster architecture. Using the second parameter causes the ClusterHeads to consume less power for communication with their neighbors. We provide the node with the capability of learning, by using Case-Based Reasoning (CBR). In accordance with the learning capability of the nodes, the Tnb and Dnb parameters are intelligently calculated and an optimum clustering algorithm is presented. The main goals of this article are increase of stability of clusters and decrease consumed power during interaction between clusterhead and its neighbors. Through simulations we have compared the performance of our algorithm with that of the Lowest-ID (LID), Highest Degree (HD) and Weighted Clustering Algorithm (WCA) algorithms in terms of number of reaffiliations and power consumed in total network. Results obtained from simulations proved that the proposed algorithm achieves the goals.

References