Personal tools
Barrage Relay Networks
| What |
|
|---|---|
| When |
May 06, 2010 from 02:00 PM to 03:00 PM |
| Where | Engr IV 67-124 Faraday Room |
| Add event to calendar |
|
Tom Halford
TrellisWare
Thursday, May 6, 2010 at 2:00pm
Engr IV 67-124 Faraday Room
Abstract
Barrage relay networks (BRNs) are mobile ad hoc networks (MANETs) based on an autonomous cooperative communications scheme that eliminates the need for link-based coordination between nodes (e.g., collision avoidance & routing). The fundamental physical layer resource in a BRN is not a point-to-point link, but rather a portion in space and time of a cooperative, multihop transport fabric. Controlled barrage regions (CBRs) define the spatial extent of multihop data transmissions in a BRN, while barrage access control (BAC) protocols coordinate the time-multiplexing of CBRs so as to avoid collisions between distinct unicast, multicast, or broadcast flows.
In the first part of this talk, the design rationale for BRNs -- i.e., the motivation for a MANET architecture targeting broadcast traffic first and foremost -- is presented. We then describe the basic building blocks of BRNs including the efficient barrage flooding protocol, as well as the aforementioned CBR and BAC concepts. Along the way, a number of new theoretical results are presented, including:
(i) A demonstration that the data rate and latency scaling for broadcast traffic afforded by BRNs is optimal. Critically, this scaling result is achieved via a practically realizable protocol that does not depend on the maintenance of a "broadcast backbone".
(ii) A study of the reliability of multihop data transmission in BRNs. A simple linear network model is used to illuminate the improvements in packet delivery ratio that result from the autonomous cooperative communications scheme employed by BRNs.
(iii) A study of the stability of multihop CBRs in BRNs. Much as route lifetimes inform Layer 3 protocol design in traditional MANET architectures, an understanding of the expected lifetime of CBRs is critical to the design of future barrage access control protocols.
We conclude this talk with a discussion of a number of open problems concerning the theoretical limits of -- and practical implementation schemes for -- barrage relay networks.
Biography
Thomas R. Halford received a Ph.D. in Electrical Engineering from the
University of Southern California. In May of 2007, he was offered a
Director's Postdoctoral Fellowship at the Los Alamos National
Laboratory, which he declined in order to join TrellisWare Technologies,
Inc. Dr. Halford's research interests include coding & modulation,
tactical mobile ad hoc networking, and situational awareness.
