The Case for Opportunistic Communication

The ability to exploit transient communication opportunities allows mobile systems to trade off higher performance and dramatically lower energy and monetary costs for an increase in algorithmic complexity and variability in message delivery delays. I argue that such a mode of communication will play a significant role in future mobile communication systems. In this talk, I will motivate the need for opportunistic communication by presenting some non-trivial application contexts. I will describe the requirements for a practical system for opportunistic communication and outline the Opportunistic Communication Management Protocol, developed over the last three years at the University of Waterloo, that meets (most of) these requirements. I will conclude with a sketch of the underlying open problems in optimization, prediction, and data privacy that still need to be solved to allow opportunistic communication to be widely deployed in future mobile systems.