Barun Saha's blog on AI and Networks

[21 March 2015] Update: The original URLs for the Software Engineering and Advanced Network Technologies Virtual Labs are temporarily not working. You can access these labs here: vlssit.iitkgp.ernet.in/isad/isad/ and  vlssit.iitkgp.ernet.in/ant/ant/ The Virtual Labs project has been formally launched today. This is a day perhaps many of us have been waiting for long. The Virtual Labs is an ambitious project undertaken by Ministry of Human Resource Development (MHRD) as part of the National Mission on Education through Information and Communication Technology (NMEICT) program. This project, I believe, has the potential to bring considerable improvement in science and engineering education in India. The world has changed, and learning (or teaching) is no more confined to the traditional books. Rapid progress in technologies and increasing accessibility to Internet gives ample scope to gain clear understanding of the subject matter. However, something still lacked there. Co...

Today, we will play around a bit with the ONE simulator, specifically with the Epidemic routing protocol. We will simulate two scenarios and look at the results. Detailed analysis of the results are left out for the time being. Parameters Here are few common parameters for the simulations. Group.movementModel = RandomWaypoint Group.msgTtl = 300 (5 hours) MovementModel.worldSize = 450, 340 Scenario.endTime = 14400 (4 hours) Stats are collected from MessageStatsReport from a single run.

A key difference that can be observed between Internet and DTN routing is that Internet routing is based on packet forwarding. Whereas in DTN, routing mostly involves message replication. A node in DTN, which wants to transfer a message to the other node in contact, creates a copy of that message and forwards that * copy * to the other node. This process is repeated at every (or limited # of) node(s) until the message reach its destination. In Internet it is only the sender who buffers the message until an acknowledgement is received from the other end of the connection. Epidemic and Spray-and-Wait are two popular DTN routing protocols using such principle. Epidemic protocol replicates the message at every node it comes in contact with. Spray-and-Wait imposes certain limit on the # of replications to be done.

How severe is the effect of congestion in case of Delay Tolerant Networks? I understand that DTN is still evolving, with very less number of real life applications. But did any of them suffer from network congestion? If so, what they did to avoid / control it?

Contents Introduction Why? Target Audience Pre-conditions The Scenario General Steps Configuration File(s) Specify Scenario Settings Specify Network Interfaces for Nodes Create a Group of Nodes Specify Motion Patterns Specify Traffic Pattern Specify Reports to be Generated Running the Simulation The ONE tutorial has moved to a new page. Please click here to follow. Introduction Opportunistic Network Environment is a Java-based simulator targeted for research in Delay Tolerant Networks (DTNs)/Opportunistic Mobile Networks (OMNs). Apart from letting a user to simulate different scenarios quickly and in a flexible manner, it also provides an easy way to generate statistics from the simulation performed. ONE can be run on Linux, Windows, or any other platform supporting Java. Details about ONE, including how to download and run, can be found from the official site . A list of common questions and their answers are listed here . Read more »