Papers and Essay's

This page contains the papers and essays I wrote as part of my post graduate course in computer science while at IIIT-Bangalore.

1. Big Kahuna - A Framework for performing massively distributed computation using grid of stateless transient nodes.

abstract

This paper explains how to achieve massively distributed computing using many stateless nodes consisting a mix of web browsers, mobile and thick clients which can perform computation on demand. This architecture solves mutually independent problems that do not require synchronization and inter node communication during computation. The arrangement uses javascript enabled browsers which act as multiple nodes connected to a web server acting as the controller. The system can scale rapidly as more browsers connect to the server and can solve a multitude of problems, especially those for which the communication overhead between partial solutions is low and there is little or no inter node communication. This approach makes the client platform independent and inexpensive to implement. The paper also talks about the advantages and disadvantages of such an approach and also how such a technology is feasible for large scale applications such as data mining, pattern recognition and many more.

Download PDF of full text. You can also see the attached presentation and video demo here

2. Beta Barrel Transmembrane Protein Prediction using TMBHunt algorithm on the Cell Broadband Engine

abstract

Beta-BarrelTransmembrane (BBTM proteins perform diverse functional roles including bacterial adhesion, structural integrity of the cell wall, and material transport. These proteins are much harder to identify due to the big variation in the short transmembrane segments. There are few methods available to identify BBTM proteins by exploring properties such as sequence profiles, beta-barrel score and signal peptides, the distribution of multiple properties on protein sequences. TMB-Hunt is one such open source method that uses a k-Nearest Neighbour (k-NN) algorithm to discriminate between BBTM and Non-BBTM proteins on the basis of their amino acid composition. In this paper, we present the implementation of TMB-hunt on Cell Broadband Engine (CBE). CBE (Cell Broadband Engine) is a heterogeneous multicore processor with one Power PC Processing Element (PPE) and eight Synergistic Processing Elements (SPE). We take advantages of CBEs parallel processing to optimize the performance of TMBhunt to achieve super linear scalaility. By parallelizing an otherwise serial operation, our implementation promises faster prediction of the BBTM/AHTM.

Download PDF of full text

3. Convergence

abstract

It’s hard to imagine a world without the internet. A world where mails would take days to reach, where information was either spread through word of mouth or was locked up in books. Telephones and more recently, mobile phones, have moved on from being luxuries to necessities. Telephones were invented more than a century ago and computers more than half a century ago. Though both these technologies seem to fit like a puzzle to give us the mighty internet, the convergence of telephony and computing together to form a global communication network was hardly a given. Voice and data do not flow separately through networks; Voice is just another type of data. In a world where it’s hard to draw the line where telephony ends and computing begins, it’s very important to know what shaped such a convergence of technology. What gave rise to the Info - Communications Industry which has this convergence of technologies at its heart. In this essay, I try to summarize my views about the convergence of telephony and Computing and also highlight essential milestones that were to shape this phenomenon.

I shall start of by highlighting the computing and software industry and then go on to telephony. The disorder in chronology is intentional as I try to relate advancements in telephony to computing which make little or no sense without knowing what was happening in parallel in the computing industry. The last section will talk about convergence of these two industries to become the internet and the network of networks.

Download PDF of full text

4. Role of Standards in the Information Economy

abstract

Standards have been pivotal in establishing the info-communications industry as we know it. The info-communication industry is regarded by many as the unexplained new economy, but deeper analysis reveals that standard economic properties, which apply to any other industry, apply to this industry as well. Almost every industry goes through the cycle of innovation, mass production, standardization and consolidation. Growth is always characterized by standardization. This is an attempt to restore stability in a chaotic market. An effort to profit from economies of scale created due to a standard.

This essay talks about the emergence of standards in the info-communications industry, the reason they come about and their importance to the customer and the supplier. The essay tries to highlight the advantages of standardizing a market, the pitfalls, the implications and the opportunities created by it. The order in chaos theory isn’t sufficient explanation to the need of standards. Standardization typically changes the markets, kills monopolies, opens up the market and fosters co-operation. This essay will take multiple viewpoints and analyze the role of standards in the network economy.

Download PDF of full text

5. Software Bottleneck

abstract

Building large software systems has been a challenge from the very beginning of software development. The emphasis is always on maintaining and optimizing cost, quality and schedule of a software project. Yet, if you were to compare the growth in productivity or priceperformance gain in software over hardware, you would see an exponential increase in the hardware side and almost a negligible rise in software. The crux of this problem is what forms the software bottleneck. The bottleneck characterizes the shortcomings in growth of software as compared to its hardware counterparts, even though they both work together in unison as a computer. Why is software building such an unpredictable activity? Its almost 40 years since the development of an independent software industry in the world, and yet, after all these years, there is still no clear solution to this bottleneck.

Promises of disruptive technologies, methodologies, tools etc have failed to deliver any significant improvement in productivity of software. Software progress is not slow; hardware progress is very fast. No other technology since civilization began has seen six orders of magnitude price-performance gain in 30 years1. Still, there are optimists that hope for a radical new technology or innovation that will deliver, and many self proclaimed evangelists in the field advocate several current day/upcoming technologies as silver bullets. A deeper look at the prediction will definitely show marketing and promotional motives in those claims.

This essay will try and highlight key aspects of the software bottleneck, its reason for existence, its manifestations, its implications and attempts at resolution. Moore’s law is being challenged as the silicon cannot get any smaller and we may have hit the optimal, but even to this day the software bottleneck exists and attempts at its resolution is only tightening the bottleneck.

Download PDF of full text

6. Barriers to Universal Access of Information on the Internet

abstract

The late 20th century is characterized by the Information and Technology Revolution. Rapid advances in communication technology have made vast amounts of information accessible. The availability of large amounts of information spawns the need for a variety of mechanisms to deliver the information. If we look at mechanisms of delivery of information in the last decade, broadcast mechanisms like the television, print media and radio have been overshadowed by the giant that we know as the Internet. The explosion in adoption rate of the internet has created a demand for faster, better and cooler products in the consumer space. Given the ease of reproducibility of concepts on the internet, most unique concepts have clones in almost every country and their time to enter the market is dramatically small. This puts pressure on the businesses to innovate fast and provide products with as many features as possible in order to provide more value to customers. A look at the internet of the late 90's, or what in popular jargon is called the Web 1.0, and the internet of today, you will see a stark difference in mechanisms of delivery, usage and interface of the website. The non uniform structuring of content, multimedia, varied formats, advertisements, clutter and other elements make the actual content of a digital resource hard to access. For people with disabilities, interfacing the computer is a non trivial task accomplished by assistive technologies and inaccessibility of information due to the above mentioned factors proves to be a far greater problem.

Computer technology and the increasing popularity of the internet have tremendous potential to broaden the lives and increase the independence of people with disabilities. But problems with accessibility of content on the internet are proving to be barriers to usage. Accessibility of information and hurdles to achieve universal access will be the focus of this essay. I shall write about the personal computer and modes of interacting with it and the wholesome experience which has problems with accessibility. The design of the modern day PC is quintessentially inefficient and leads to a problems with adoption, usability and accessibility. I shall also write about efforts to provide universal access to information and its pros and cons. The new wave of the web brought along with it news paradigms of interaction with the web. People could not only consume, but also contribute content. Faster and cheaper bandwidth has increased the prominence of multimedia and dynamic content on the web and as this half baked technology finds its way onto the internet, regular rules of accessibility, that applied to textual content in general, is fast becoming obsolete. I shall try to highlight some prominent changes to the paradigm of the web and also how these changes affected accessibility. I shall also talk about standards for accessibility of information on the web, its evolution and its confusing role as antidote to the problem of accessibility and also how inaccessible websites are socially excluding the very people for whom information is truly empowering.

Download PDF of full text