[Nrg-l] Remainder - November 5 @ 3pm

Jorge Londoño jmlon at cs.bu.edu
Sun Nov 4 20:51:54 EST 2007

This Monday we will have two work-in-progress presentations:

Sowmya Manjanatha
Active Location Verification Using RF Jamming
Radio Frequency (RF) interference is generally considered a nuisance in 
wireless networks. RF jamming is an interference generated by an 
``attacker'' that causes a genuine wireless signal to decrease its 
Signal-to-Noise ratio almost canceling at the receiver.  In this paper, 
we argue  RF jamming could be used beneficially to solve the problem of 
in-region location verification.  Specifically, we exploit the property 
that signals can only be jammed within the intersection of the ranges of 
multiple transmitters, and propose a solution based on a framework that 
strategically superimposes multiple in-phase RF signals over a space -- 
a process we call {\em healthy jamming}.  The jammers are deployed such 
that the area to be protected forms intersection regions. The 
verification process involves generating jammed signals and verifying 
that a node claiming to be in the non-intersection region correctly 
recognizes a clean (not affected by interference) signal.  Our design 
approach leverages the unique characteristics of signal interference to 
reliably determine the trust-worthiness of a node. Our approach has the 
advantage that the provers are passive listeners as opposed to the 
currently prevailing
philosophy of provers being active.  In doing so, our approach naturally 
eliminates the threats caused by provers with powerful equipment such as 
directional antennas. We verify robustness and determine the performance 
bounds of the proposed solution through detailed analysis.

Marisa Affonso Vasconcelos
Extracting Location from Contact Traces

Localization of nodes is essential for many applications. Sensor 
networks, routing in mobile environments as well as caching can benefit 
by having this extra information. Applications dealing with emergencies 
and rescue operations need precise and accurate localization 
information.   Many node localization techniques have been proposed, but 
most of them lack in some feature (non-convex areas, non symmetric 
communication, other types of constraints, etc) that will be desirable 
to have.   For example, some of the techniques requires special hardware 
or are only effective when the nodes are close to nodes whose location 
is known a priori. Sextant is one such solution which seeks to 
incorporate most of the desirable features. It is a constraint --based 
framework which relies on connectivity information.

The novel aspect of Sextant is that it relies on positive as well as 
negative spatial constraints. Positive constraints are based on the 
reception of a beacon from a nearby node. If a node cannot receive 
direct transmissions from another node, this constitutes a negative 
constraint. However, Sextant and the other techniques do not give us 
accurate localization estimates if we incorporate mobility of nodes. 
Although, mobility brings us several issues, we believe that mobility 
can offer us extra information to improve our estimations.  In this 
talk, I hope to review Sextant as well as recent work we have been 
pursuing in incorporating mobility.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cs-mailman.bu.edu/pipermail/nrg-l/attachments/20071104/2d235c86/attachment.html 

More information about the Nrg-l mailing list