Student
Student |
|
| Anuj Agrawal | ||
|
Senior M. Tech. Student
K. R. School of Information Technology, Indian Institute of Technology Bombay, Powai, Mumbai - 400 076 (India). Phone: +91-22-2576 4955 Fax: +91-22-2572 0022 
|
|
An advanced networking course which deals with the Quality of Service (QoS) aspect of networks. Probably it was the first time that students had to write a term paper in this course instead of working on an assignment. Objective? As they say, to improve our technical skills. But how does one improve her technical skills if she does not get any kind of feedback (apart from marks) from the instructor? I would say, a wasted 15 % weightage. Rest all was good. A project on implementing an MPLS network was fun. Quizzes and exams were designed meticulously. Overall, an interesting course for networking guys.
Have you ever wondered how applications run in a distributed environment? If yes, then figure it out in this course. With several quizzes and exams you would never forget the concepts. A term paper with the objective to look at some aspect of distributed systems more carefully. But, I guess, an assignment on CORBA or RMI would have been better instead. The worst part of this course is the monotonicity in the instructor's voice. But never mind. The interesting concepts will not let you sleep so easily. Overall, another interesting course. But beware, grading can be strict.
I guess, I would not be wrong if I name this course as probability in networks. Yes, the course deals with how probability plays an important role while designing a network. Guess what would happen if the blocking probability of a network is very high at some given load. There were two unamusing small write-up assignments. A small simulation assignment to provide better insight of the concepts. And an assignment on Linear Programming (LP) probably meant to make us learn how to use an LP-solver. Above all this, a course project where I ran some more simulations. Overall, a heavy course. Probably I was right about taking only one course this sem and play it cool.
A mandatory course designed to achieve some unclear and undesirable objectives. The aim of this course is to give students an idea of how to work as a team. With this objective in mind, a problem statement was picked - to develop a plugin for eclise which would enable coders to build software architecture. The whole class of 46 was divided into few teams, each team being responsible for handling some part of the problem. End Results: A plugin was made but none of the students learnt anything. Thank God, the course has been scrapped now!
This comes from a bunch of institute electives that were offered. Yeah, a few courses from other departments are labeled as institute electives and it is mandatory to take at least one such course. Objective? I dont know. I chose this one as it is more or less based on data mining. The course is carefully designed and the lectures were lucid so that even people from non-CS/IT background did not find the concepts difficult to comprehend. Overall, a good course.
A course that covers the software aspect of embedded systems. An assignment to make a machine that could follow another machine with a changing velocity profile was the key component of the course. I wish if design part of the machine could also be included! It covers a range of interesting topics like resource management and scheduling paradigms for real systems. But something was missing. You cannot teach things like Esterel and Handel-C without an appropriate assignment. Overall, not a course that could motivate me to further work in this area.
Many a times, people get attracted to learn some new things. Things that attract or intrigue them in a glamorous way. The same happened to me when I chose this course. It was fun to learn how graphics rule our world. An assignment on OpenGL to provide some insight into graphics programming. Another one on POV-Ray to probably bring out some creativity along with providing insight into ray tracing. A presentation of some paper to get a better picture if a topic not covered in the class. A good instructor who develops every concept from the fundamentals so that people (like me) who have not taken a basic course on graphics can also grasp. The best part was to know how special effects are created in movies! The worst part was, as usual, strict grading. :(
A mandatory course that aims at the impossible: to improve our communication skills. I call it impossible because almost all the students find themselves yawning or dozing during the lectures. Or, at least the back-benchers do. It is worthless to have such a course mandatory which neither creates interest nor does it succeed in its aim (in case of most of the people).
Commonly known as F-Lab the course aims at improving our programming skills and habits, and learning some new tools. At the time of registration, I knew preliminary programming only in a few languages viz. C, C++ and VB, and Windows was the only operating system I was familiar with. But now I can boast of having flexed my analytic muscles on Unix Shell Scripting, CGI, PHP, Unix System Programming, Java, JavaCC, Perl, and much more. Really too much to boast about! :D But every damn thing has a bad part to it. The FLab assignments have killed my passion of programming. It really SUCKS!!!
This is a course that was most loved by me (and perhaps all those who love Mathematics, especially the probability and statistics part of it). Its a course with lots of potential but all wasted by the course instructor who dragged the course as fast as he could. And, above all, my bad luck that I missed a few classes of it all of which happened to had surprise quizzes which led me to screw up with my grades in this course.
This course is designed to give an insight into the mobile world. It is quite worth to know mobility, which eases our lives, creates huge and complex problems for the researchers and how these problems are tackled. A hands-on experience with network simulators like OpNet and QualNet plus a short term project in J2ME act as spicy flavour to the course. Anybody with interest in networking would like it. And, the way of teaching of the instructor further inflates curiosity and appreciation of concepts.
The course explains the basic fundamentals of computer networking, particularly the Internet. Almost all of the protocols used today in the Internet are covered. Plus, an assignment of Network Programming makes things even clearer from a developer's viewpoint. Cool nature of the instructor kept me relaxed and helped me have a better understanding of the concepts. Though my grade in this course was not what it should have been but I learned a lot. A person like me who had not done his B. Tech. in Computer Science or Information Technology must take this course to get an insight as to how the Internet works today.
A. Gumaste, D. Diwakar, A. Agrawal, A. Lodha, N. Ghani, "Light-Mesh: A Pragmatic Optical Access Network Architecture for IP-Centric Service Oriented Communication," invited paper, Elsevier's Journal of Optical Switching and Networking, Vol 5, Issues 2-3, pp 63-74, Jun 2008
Abstract: Contemporary deployments of optical access networks are based on the principles of Passive Optical Networks (PONs). PONs deploy a star topology and dual wavelength for communication between the center and ends of a star. The star topology requires that each end-user be connected to the star splitter (usually a passive coupler). We argue in this paper that while adhering to the requirements of access networks, we are able to provide a better topological solution in terms of the cost-factor and the ability to upgrade to a greater bandwidth. This solution, called a light-mesh, is based on the concept of pragmatic optical packet transport or light-frames results in a unique node architecture, interconnection matrix, and communication protocols. We begin by investigating into the node architecture that is required for a mesh network in the access area. The proposed node architecture has unique benefits in terms of being able to support the intermittent communication in the access area − nodes are not always powered ON, despite which, it is important to maintain mesh connectivity. Hence we propose the use of largely passive components in node architecture design. Passive components in a mesh lead to collisions of packets in the access area, for which we propose a unique collision detection and recovery scheme based on a logical time-overlap method. Collisions make the end-to-end delay uncertain. Analysis of the associated delay is performed. We then propose algorithms to build such a light-mesh network. These algorithms are investigated in terms of network built-out costs and these costs are compared to a PON topology. Cost differences and a performance comparison with PON are presented as part of the numerical analysis.
A. Gumaste, N. Ghani, P. Bafna, A. Lodha, A. Agrawal, T. Das, J. Wang, "DynaSPOT: Dynamic Services Provisioned Optical Transport Test-Bed - Achieving Multirate Multiservice Dynamic Provisioning Using Strongly Connected Light-Trail (SLiT) Technology," IEEE/OSA Journal of Lightwave Technology, Vol 26, Issue 1, pp 183-195, Jan 2008
Abstract: In this paper, we report on the dynamic services provisioned optical transport (DynaSPOT) test-bed − a next-generation metro ring architecture that facilitates provisioning of emerging services such as Triple Play, Video-on-Demand (VoD), pseudowire edge-to-edge emulation (PWE3), IPTV, and Data Center Storage traffic. The test-bed is based on the recently proposed strongly connected light-trail (SLiT) technology that enables the triple features of dynamic provisioning, spatial subwavelength grooming and optical multicasting − that are quintessential for provisioning of the aforementioned emerging services. SLiT technology entails the use of a bidirectional optical wavelength bus that is time-shared by nodes through an out-of-band control channel. To do so, the nodes in a SLiT exhibit architectural properties that facilitate bus function. These properties at the network side include ability to support the dual signal flow of drop and continue as well as passive add, while at the client side include the ability to store data in order to support time-shared access. The latter (client side) improvisation is done through a new type of transponder card − called the trailponder that provides for (electronic) storage of data and fast transmission (burst-mode) onto the SLiT. Further in order to efficiently provision services over the SLiT, there is a need for an efficient algorithm that facilitates meeting of service requirements. To meet service requirements we propose a dynamic bandwidth allocation algorithm that allocates data time-slots to nodes based on a valuation method. The valuation method is principally based on an auctioning scheme whereby nodes send their valuations (bids) and a controller node responds to bids by sending a grant message. The auctioning occurs in the control layer, out-of-band and ahead in time. The novelty of the algorithm is the ability to take into consideration the dual service requirements of bandwidth request, as well as delay sensitivity. At the hardware level, implementation is complex − as our trailponders are layer-2 devices that have limited service differentiation capability. Here, we propose a dual VLAN tag and GFP-based unique approach that is used for providing service differentiation at layer-2. Another innovation in our test-bed is the ability to support multispeed traffic. While some nodes function at 1 Gb/s, and others function at 2.5 Gb/s (using corresponding receivers), a select few nodes can support both 1- and 2.5-Gb/s operation. This novel multispeed support coalesced with the formerly mentioned multiservice support is a much needed boost for services in the metro networks. We showcase the test-bed and associated results, as well as descriptions of hardware subsystems.
A. Agrawal, A. Gumaste, M. Chamania, N. Ghani, "Topology Abstraction Algorithms for Light-Mesh: An Alternate Model for PON," Proceedings of IEEE/OSA 24th Optical Fiber Communication (OFC) Conference, San Diego, USA, Feb 2008
Abstract: Light-mesh − an alternate solution for access networks is presented. Two heuristic topology algorithms are discussed and simulated showing cost and performance benefits.
My personal page is under construction. To find my contact details click here. To read my résumé click here.
| IIT Bombay | KReSIT | People at KReSIT | Faculty |