Informatics Publishing Limited

M. Dheenadhayalan ¹, J. Mannar Mannan ¹, M. Sundarambal ²

Affiliations
1 Department of Information Technology, Anna University of Technology, Coimbatore, Tamilnadu, IN
2 Department of Electrical and Electronics Engineering, Coimbatore Institute of Technology, Coimbatore, Tamilnadu, IN

Abstract

Wireless ad hoc networks susceptible to intrusions as they operate in an open medium and use co-operative strategies for network communications. Solutions that are designed for wired networks are not always suitable for wireless networks, especially Mobile Ad Hoc Networks (MANET) are in Dynamic nature. To obtain an acceptable level of security for MANET’s, Traditional security solutions like Agent (mobile agent) based are combined with intrusion detection mechanisms. One method is to have IDS is running on every mobile host in a network, which runs a local detection engine analyzing local data for anomalies. A cooperative detection mechanism decides whether there is an intrusion, with all nodes taking part in the decision process related agent. But MANET nodes typically limited battery power. Thus it is not efficient to make each MANET node always a monitoring node, especially when the threat level is low. Instead a cluster of neighboring MANET nodes can randomly and fairly elect a monitoring node use agent and the cluster head. If a compromised node happens to be selected as the cluster head, it can launch attacks without being detected. The cluster head selection based on the fitness function based. The higher fitness function value take the monitoring the nodes of others in the cluster network.

Keywords

Mobile Ad Hoc Network (MANET), Cluster Head, Mobile Agent, and Fitness Function, Intrusion Detection..

Full Text

     

M. Dhivya ¹, D. Shyam ¹, M. Sundarambal ²

Affiliations
1 Department of Electrical and Electronics, Anna University of Technology, Coimbatore-641047, Tamilnadu, IN
2 Department of Electrical and Electronics Engineering, Coimbatore Institute of Technology, Coimbatore-641014, Tamilnadu, IN

Abstract

Clustering approach is a widely used technique in wireless sensor networks to exchange information between various nodes, for enhancing the performance and energy efficiency. In this paper, a Timeline approach for energy efficient cluster formation with Hybrid Optimization encompassing Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) is proposed. Timeline Cluster formation Algorithm (TLCA) incorporates multi-individual metrics in cluster formation and energy efficient optimization technique for multiple sensor nodes. The proposed technique is to implement time based elemental approach for each protocol in which each node makes information based on local and global decisions. The performance of the proposed method is compared with other clustering protocols with respect to energy level consumed and network life time. The elicit of the proposed technique is demonstrated with simulation results. The simulation results show that TLCA can produce time based network topology and reduce network cluster size.

Keywords

Clustering, Hybrid Optimization, Energy Efficiency, Timeline Algorithm, Pheromone Updating Rule, Network Lifetime.

Full Text