Q-Analyze Tool to Detect Malicious and Black Hole Nodes in NS2 Simulation for AODV

Abstract

Network simulator (NS2) was developed by UC Berkeley as an open source Internet Protocol Networks simulator. It uses TCL for front-end interpretation based on C++ back-end engine. TCL based script used in creating the network scenario producing an ASCII formatted trace file capturing all events during the simulation. The produced trace file captures all necessary information for studying and assessing the network performance like source to destination packets transfer, packets loss, dropped packets and so on. But, the resulting trace file is just a chunk of ASCII information in a textual file format and rather difficult to assess and measure with the currently available post processing techniques. The trace file might also be huge in size in some cases making it not so easy to parse for meaningful information. Q-Analyze is an NS2 trace file analysis tool proposed to simplify the process of data extraction for performance metrics measurement and producing an easy to understand network simulation report. This software is basically a tool for mining NS2 trace files for the network simulation environment of NS2. Q-Analyze tool involves three layers. First layer is the source data layer where the NS2 trace file data is extracted in row format. Second layer is the data processing layer in which row data acquired from the source trace file is processed and transformed into meaningful format to be consumed in third layer. Presentation layer is the third layer where meaningful data is prepared in a tabular format for network performance study. Different performance metrics like packet delivery fraction, throughput, and end-o-end delay are produced in a user friendly report. All malicious and black hole nodes are captured and reported in a separate report showing their behavior and why they are reported as misbehaving nodes. Through Q-Analyze tool users can do network scenario performance study using an easy to use GUI allowing them to concentrate on algorithms development instead of spending much time on data post processing and performance metrics calculation and presentation.