Improving AODV using Acknowledgment based Punishment Algorithm for MANETs in the Presence of Selfish Nodes

Abstract

Link breakage is a routing challenge in Mobile Ad-hoc Network (MANET) which affects the performance of Ad-hoc On-demand Distance Vector (AODV) routing protocol. Several reasons are stated for link breakage, but one of the major challenges is the presence of Selfish Nodes (SLFNs). In AODV nodes are assumed as a Normal Node (NN), but some nodes may behave selfishly to conserve their resources such as power and memory. This selfish behavior of nodes leads to refuse participation and drop packet which issued from another neighbor node. This selfishness misbehavior leads to several impacts as such; network partitioning, refusal to accept requests, and packet drop. To resolve the above issues, in this research Acknowledgement based Punishment on AODV (Ack-AODV) algorithm is designed. The proposed Ack-AODV algorithm identifies SLFN and punishes them to be triggered and become a NN in the network. Based on node residual energy and their request replay credit the nodes which drop a packet and have low credit value are identified as SLFN and will be blocked from any service. By applying a punishment mechanism those SLFN that are blocked will be demanded to cooperate. If SLFN reply for the acknowledgment request sent from an authorized NN and earns a credit will be revived and free from the blocked state. To analyze the performance of the proposed Ack-AODV algorithm ns2 network simulator tool is used. Based on the generating result Ack-AODV performance is measured by comparing with existing AODV. As the simulation, the result indicates Ack-AODV has achieved a good performance compare to the existing AODV. By deploying a total of 50 nodes the network throughput is increased from 80.7% to 86.3%, however end-to-end delay and packet drop rate is reduced from 35% to 23.3% and 34.39% to 25.78% respectively. To conclude, based on the comparison result Ack-AODV has a good performance in the above-mentioned metrics compared with existing AODV in the presence of SLFN. `