چكيده به لاتين
Abstract:
By increasing the use of mobile devices such as smart phones, portable computers and laptops, the mobile ad-hoc networks have a significant place in the information technology. The designing of mobile ad-hoc networks has been confronted by various constraints, such as dynamic topology, the stored energy of the nodes, huge networks, nodes mobility, service quality and physical security. Routing in such networks is difficult due to these limitations. Routing protocols in ad-hoc networks are various and vast in which divided to some categories according to nodes organization, management and methods of routing. The AODV is one of most useful on demand protocols. In the case of increasing network traffic, routing load will be extremely increased. The purpose of this thesis is to decrease the overhead of routing, considering the nodes resources and use them for routing. Therefore, in this approach, we will combine the proactive and reactive routing properties in order to use of both approach advantages together. In the proposed method, we will use the method of nodes clustering in networks. The aim is to determine a neighborhood in which routes exist proactively. Routing among these neighborhoods will be based on the requirements. In this case, the broadcasting messages for routing will be changed to multicast messages and consequently overhead is decreased. The proposed idea is to choose a node as a cluster head and the other nodes will be members of that. In this case, we have considered the clusters in two hops. The challenge is to choose the cluster head accurately. In order to determine the cluster head, it is defined as a criterion for eligibility evaluation due to the nodes resources. In this weighted definition, it has specified three features as process throughput, remained energy and the amount of memory to calculate the eligibility of every node. Each node calculates the eligibility of itself and then exchange it to the rest of two hop neighbors. In this way, the node with most resources will be selected as a cluster head. In the proposed method, the effort was to decrease AODV reactive protocol disadvantages through the proper use of advantages of proactive and reactive protocols and in return, the advantages of proactive routing are used. In addition, the proposed clustering algorithm will decrease difficulties of scalability and the overhead of ad-hoc networks. In this method, it has been avoided from changing the cluster head and efficiency of the routing in ad-hoc networks will be improved.
Keywords: mobile ad-hoc networks, proactive reactive routing, AODV protocol, clustering.
