> Projects

DataWiz App

We've developed a personalized bandwidth manager for iOS and Android that helps manage usage on end users' mobile devices.  The DataWiz app predicts future usage and gives daily recommendations to make sure that users stay within their monthly data caps.  DataWiz also educates users, breaking down usage by time and location and (on the Android platform) by individual application.

TUBE (Time-Dependent Pricing)

We've implemented an end-to-end system for offering day-ahead time-dependent pricing to broadband users.  The basic idea is to offer lower prices in less congested periods, encouraging users to shift some of their traffic from congested to less congested periods, thus relieving the peak load on ISP networks.   Our system computes the optimal prices to offer, given estimates of user reactions to these prices.   ISPs can then monitor the resulting usage pattern over the next day and adapt the next day's prices to any observed changes in user behavior.

IDS (Intelligent Demand Shaping)

We study the effects of allocating available bandwidth to base station users in a fair and efficient manner.  In times of network congestion, ISPs can utilize usage history data in a Nash bargaining framework to decide how much bandwidth to give to each user.  Lighter users, whose behavior does not stress the network, are given "priority" in receiving bandwidth, while heavier users may be "punished" for contributing more to the network load.

MAP (Measurement, Analytics, and Profiling)

We attempt to understand users' behavior through an analysis of temporal and application-based usage patterns.  We collect individual usage data on ISP networks and use it to predict future usage, break down usage by location, and cluster users by the similarity in their temporal and application usage patterns.  Our analysis thus studies usage predictability on the network and individual user levels, as well as similarities in usage patterns across different users.