Another Look at Dynamic Ad-Hoc Wireless Networks

Abstract

The price, performance, and form factors of sensors, processors, storage elements, and radios today are enabling the development of network-supported applications for a wide range of environments, including the monitoring of disruptive phenomena, object tracking, establishment of on-demand network infrastructure for disaster relief or military purposes, and peer-to-peer vehicular or interpersonal networks. However,while in theory ad hoc networks are the ideal vehicle for such applications, the practice today is far from this theory.In this talk, I argue that many of the limitations of ad hoc networks today stem from the fact that the architectures and protocols used for them are in many ways a derivative of the Internet architecture, and describe a research agenda that considers developing ad hoc networks without having to adhere to many of the design choices that,until now, have proven so successful for internetworking of wired networks.

This work was supported in part by the Palo Alto Research Center and by the Baskin Chair of Computer Engineering at University of California, Santa Cruz