After my wonderful years at Penn, I am finishing my PhD in June 2015, and joining EXL afterwards.
This page, therefore, will no longer be maintained.

Please visit my new personal webpage at
Home    Publications    Links    Resources
Image: Mohammad Hadi Afrasiabi

Mohammad Hadi Afrasiabi

(Hadi Afrasiabi)

PhD Candidate,
Electrical and Systems Engineering,
University of Pennsylvania

Lab 306, Moore Bldg
200 South 33rd Street
Philadelphia, PA 19104
Phone: (215) 628-4629

Email: afram (the at sign) seas (dot) upenn (dot) edu

I received my BSc in Electrical Engineering, Communications, from Sharif University of Technology in 2009. In the same year I started my PhD at the University of Pennsylvania and joined the Multimedia and Networking Lab. Here I am working with Professor Roch Guerin .

My research is in the broad area of Social and Technological Networks and Network Economics. Examples include modeling and analysis of opinion spread in a networked environment, adoption patterns of technologies and service profit maximization.

The first research problem I worked on was modeling the adoption patterns of a "User-Provided Connectivity" service, where a user shares his/her home connectivity with the "roaming" users that are in the vicinity. Such a service provides an alternative for the traditional infrastructure-based connectivity and, in contrast to that, realizes connectivity organically as more users join the service. Our model captures both positive and negative externalities that such a system exhibits. The equilibria of this system were identified and a two-price semi-optimal pricing strategy was proposed.

Another research problem that I worked on is about deploying variations of the traditional Ising spin glass model to capture the spread and formation of opinions in a networked environment. I have been looking at various variations of this classical model, including one in which party affiliations of the users are captured by the model. Another variation is one in which the influence of one user on the other is determined based on the similarity of their "profiles" of personal characteristics. We have determine conditions under which the modeled system converges to different of equilibria.