Detecting the intruder actions at the application and routing protocol level is the main problem considered in this project. To identify the intruder actions a simple OPNET modeler simulation is done against the MANET network and the corresponding scenarios are created and computed to estimate the intruder actions on the MANETs.

A small scale MANET is created for the simulation purpose with 20 mobile nodes and an individual wireless LAN server with AODV as the routing protocol to initiate the communication patterns across the network.  Intruder actions are detected based on their behavior on the routing protocol and the application and for this three scenarios are created where these scenarios vary at the node level settings of the AODV routing protocol.

In the first scenario all the nodes are set against the default AODV parameters and in the second scenario two nodes like node 5 and node 10 are set with the varied AODV parameters such that they act as the attack nodes and send the intruder traffic across the network. In the third scenario node 15 is also identified as the attack nodes and the corresponding routing parameters of AODV are changed such that even this node acts as the attack node and thus the overall performance of the network due to the attack nodes is estimated using the simulation. 

All the scenarios are run for 2 minutes against the application and routing protocol metrics and in this simulation a simple IP telephony voice application is used and the results are compared and evaluated to understand the role of intruders in generating the required traffic and packet manipulations across the network. The key aims and objectives identified for this project are as given below.

Paper Written & Submitted by Harsha