What do I do for a living?...

Research: Stingray | JAVeLEN | ADROIT

From July 2003 to August 2004, I worked as a Postdoctoral Researcher at ICIR, the ICSI Center for Internet Research.

Research: Exploring the evolution of both the Internet's most heavily used transport protocol, TCP, and the current network environment with respect to how the network's evolution ultimately impacts end-to-end protocols. The traditional end-to-end assumptions about the Internet are increasingly challenged by the introduction of intermediary network elements (middleboxes) that intentionally or unintentionally prevent or alter the behavior of end-to-end communications. Our research provides measurement results showing the impact of the current network environment on a number of traditional and proposed protocol mechanisms (e.g., Path MTU Discovery, Explicit Congestion Notification, etc.). In addition, we investigate the prevalence and correctness of implementations using proposed TCP algorithmic and protocol changes (e.g., selective acknowledgment-based loss recovery, congestion window growth based on byte counting, etc.). We present results of measurements taken using an active measurement framework to study web servers and a passive measurement survey of clients accessing information from our web server. We analyze our results in the context of gaining further understanding of the differences between the behavior of the Internet in theory versus the behavior we observed through measurements. In addition, these measurements can be used to guide the definition of more realistic Internet modeling scenarios.

From June 2001 to December 2002 I was a inter at Sprint Advanced Technologies Laboratory in Burlingame, California. I worked in Christophe Diot's IP group doing research on traffic matrix estimation.

Research: Any Internet Service Provider (ISP) such as Sprint owns a set of Points of Presence (PoPs) in cities across the geographical area covered by its data communication network. A PoP is a set of routers co-located in a single place (e.g. a building). Each POP received an aggregated of data coming from customers located in the corresponding city. In addition, a PoP may receive data coming from peering connections to other ISPs. Many traffic engineering and performance management tasks performed by an ISP benefit directly or indirectly from knowledge regarding the amount of data exchanged by any pair of PoPs in the network. However, obtaining such information is extremely challenging because of the very large amount of data that traverses each PoP. My research at Sprint focused on alternative techniques to enable network operators to gather the necessary traffic exchange information without having to directly measuring it. My research led to the proposal of an approximation mechanism based on Economic Consumer Theory, in which the traffic exchanged between PoPs is modeled in terms of the engineering characteristics of the participating PoPs. Using this approach, only partial data needs to be directly measured from the network and the rest is filled using appropriate statistical methods. My Ph.D. dissertation thesis was the result of this work.

From July 1999 to May 2003, I was a Ph.D. student in the Computer Science at Boston University . My areas of interest are performance analysis and modelling, wireless communications, topology generation, among others. At BU I worked Professor Ibrahim Matta.

Research:My research work at the Computer Science department of Boston University centered on the development of models and software tools to generate representative synthetic Internet topologies to be used in Internet studies and simulations. The main outcome of this research was the development of a universal topology generator, called BRITE (see Software Section below).

From June 1997 until July 1998 I spent a wonderful year as a Visiting Scholar within the Supercomputing Technologies Group at the MIT Laboratory of Computer Science , under the Supervision of Prof. Charles E. Leiserson .

Research: Cilk is a language for multithreaded parallel programming based on ANSI C designed and developed at the Supercomputing Technologies group at MIT Laboratory for Computer Sciences. Cilk is designed for general-purpose parallel programming, but it is especially effective for exploiting dynamic, highly asynchronous parallelism, which can be difficult to write in data-parallel or message-passing style. My research on Cilk centered around integrating the capabilities of the language with existing mathematical tools such as Matlab in order to make asynchronous paralleling processing capabilities offered by Cilk available to these tools.

From January 1996 to July 1997 I was a Master student in the College of Computer Science at Northeastern University

Research: Worked in Scientific computing. This area research is concerned with developing and implementating new computational techniques to be applied to different scientific areas. Specifically, I worked in the research group of Prof. Bryant. W. York, looking for new alternatives regarding efficient implementation of Fourier Transform algorithms.

Return to Home Page