An enhanced cluster head selection algorithm for routing in mobile AD-HOC network

Abdulsaheb, Ghaida Muttasher (2017) An enhanced cluster head selection algorithm for routing in mobile AD-HOC network. PhD thesis, Universiti Malaysia Pahang (Contributors, Thesis advisor: Sulaiman, Norrozila).

An enhanced cluster head selection algorithm for routing in mobile AD-HOC network.pdf

Download (1MB) | Preview


Mobile Ad hoc networks have become one of the most important kinds of network because of their simple construction, which does not require any pre-fixed infrastructure. Clustering is used to reduce connection load and eliminate the routing traffic that occurs during the discovery route process. The characteristics of MANET have led to many difficulties in partitioning the main networks to a cluster, selecting a node to be the cluster head, and choosing the most suitable route for sending data from a sender to a receiver. This thesis proposes an algorithm for increasing the stability of the cluster by selecting the most stable cluster head, maintaining the cluster structure with minimum maintenance overhead, and finding the best performance routing algorithm for use over MANET. This thesis proposes a cluster based routing protocol, the Enhanced Cluster Routing Protocol (ECRP), which uses a modified cluster formation algorithm to build the cluster structure and select one of the nodes to be the cluster head. The selection of the cluster head is made based on different parameters, which are: storage capacity, load distribution, accumulative time, available power, and node movement. The scale of each node is calculated according to these parameters and the node that contains the highest parameter values is chosen to be the cluster head, which is used to decide the route and the management process for all other nodes in the cluster. An algorithm for cluster maintenance is proposed that categorizes and recovers errors according to their type. The cluster maintenance stage ensures proper delivery when sending packets, which covers link failure, node movement, cluster head movement, ordinary nodes becoming the cluster head, two nodes needing to be the cluster head, and node shutdown. The routing discovery process is divided into two stages: intra-cluster routing discovery and inter-cluster routing discovery. Intra cluster routing is defined as routing within the same cluster. Inter-cluster routing corresponds to routing between two clusters using a Global Positioning System (GPS). A model for the Ad hoc network was simulated with various cluster scenarios. The effect of changes in movement (speed) and node density (number of nodes) on the performance of the proposed routing protocols was measured. The performance of ECRP algorithms was compared with other cluster based algorithms. The results show that the ECRP algorithms is more stable and effective than existing solutions as shown in the final trust scores for throughput, packet delivery ratio, end to end delay, number of dropped packets, and normalized control overhead. The simulation results show that the ECRP increases throughput and the packet delivery ratio. The end to end delay, the number of dropped packets, and normalized control overhead was reduced.The results presented in this thesis are significant in terms of making clustering algorithms acceptable to users and improving the performance of cluster formations, maintenance and routing algorithms over MANET. Results indicate that the ECRP can achieve at least a 5% to 10% improvement compared with other cluster based algorithms.

Item Type: Thesis (PhD)
Additional Information: Thesis (Doctor of Philosophy in Computer Science) -- Universiti Malaysia Pahang – 2017, SV: DR.NORROZILA SULAIMAN, NO CD: 10790
Uncontrolled Keywords: Mobile Ad hoc networks
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Faculty/Division: Faculty of Computer System And Software Engineering
Depositing User: Ms. Nurezzatul Akmal Salleh
Date Deposited: 18 Jul 2017 08:02
Last Modified: 02 Mar 2023 04:16
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item