Load-balancing algorithm to enhance the performance of Mobile Ad Hoc Networks

Abstract

A Mobile Ad Hoc Network (MANETs) is a network comprising mobile nodes formed without the support of any centralized physical infrastructures. MANETs become a significant communication technology in modern life due to their flexibility. Scalability is of central concern among researcher community and an issue that has crucial impact on capability and capacity of MANETs. Clustering has continued to be a promising approach for MANETs because it enables efficient routing protocols and achieves high energy efficiency in high density MANETs. Clustering by definition, refers to the virtual splitting of the mobile nodes into different groups. Each cluster is managed by a cluster head, which is among the nodes within the geographical area of the cluster. The responsibilities of the cluster head include routing, allocation and coordination of resources among the cluster nodes. However, clustering faces many challenges and difficulties that require further efforts for them to be addressed. Chief among these challenges include mobility feature which leads to dynamic MANET topology, limited bandwidth and limited battery power in wireless medium. Thus, it is important to design an efficient clustering algorithm to support the robustness of the network as a result of topological modifications triggered by nodes movement in large networks with different communication workloads. In this study, the problem of load balancing in clustering algorithms is investigated in MANETs. In particular, a new Cluster head Load Balancing Algorithm (CLBA) has been developed. The main objective of this study is an improvement in a well-known Weighted Clustering Algorithm (WCA) for MANETs. The main contribution of this study is to examine the variance between communication workload of every mobile node when load balancing is performed between the cluster heads. Furthermore, cluster heads’ variable transmission range is utilized to prevent the network from re-clustering. This is intended to lessen the computation and communication overheads and optimizing the usage of the limited bandwidth in the wireless network. Simulation model is used to demonstrate the efficacy of the proposed algorithm improve the network stability by using load balancing algorithm between cluster heads. The performance of the proposed CLBA algorithm is appraised by developing an extensive simulation by utilizing the OMNeT++ software. The proposed CLBA algorithm’s performance are measured in terms of overhead clustering, cluster head lifetime, route discovery time and throughput in MANET. The results obtained are compared with the well-known Weighted Clustering Algorithm (WCA) often utilized in MANETs. The CLBA considerably out performs WCA in most of the studies’ metrics and scenarios.