[Colloq] Hiring Talk - Friday, April 8, Alberto Cerpa, UCLA

[Rachel Kalweit]

College of Computer and Information Science Colloquium

Presents:
Alberto Cerpa
UCLA

Who will speak on:
Low-Power Wireless Links Properties: Modeling and Applications in Sensor 
Networks.

Friday, April 8, 2004
2:00pm
366 West Village H
Northeastern University

Abstract:
Advances in low-cost, low-power micro-sensor and radio design have led 
to active research in large-scale networks of small, wireless, low-power 
sensors and actuators.  Radio communication is one of the critical 
components in any wireless distributed systems, but wireless sensor 
networks make extensive use of communications in order to perform the 
coordinated sensing tasks.  These systems will be deployed in many 
environments the present very harsh conditions for wireless 
communication using low-power radios, including multipath/fading 
effects, reflections from obstacles, and attenuation from foliage. 
Moreover, recently several landmark wireless sensor network deployment 
studies clearly demonstrated a large discrepancy between experimentally 
observed communication properties and properties produced by widely used 
simulation models.

New approaches and models are required to deal with vagaries of the 
communication channel.  In the first part of this talk, I motivate the 
need for precise wireless link characterization and describe the 
methodology used to collect extensive experimental data traces, using 
different hardware, in different environment, under systematic varied 
conditions. I perform statistical analysis on the data and provide sound 
foundations for our conclusions by extracting relationships between 
location (e.g distance) and communication properties (e.g. reception 
rate) using non-parametric statistical techniques.  The objective is to 
provide a probability density function that completely characterizes the 
relationship. Furthermore, I study individual link properties and their 
correlation with respect to common transmitters, receivers and 
geometrical location.  Using the modeled communication properties, we 
develop a series of wireless network models that produce networks of 
arbitrary sizes with realistic properties.  I use an iterative 
improvement-based optimization procedure to generate network instances 
that are statistically similar to empirically observed networks.  I 
evaluate the accuracy of the conclusions using our models on a set of 
standard communication tasks, like connectivity maintenance and routing.

In the second part of this talk, I describe more deeply the temporal 
properties of links in low power wireless communications.  I investigate 
short term temporal issues, like lagged autocorrelation of individual 
links, lagged correlation of reverse links, and consecutive same path 
links.  I also study long term temporal aspects, gaining insight on the 
length of time the channel needs to be measured and how often we should 
update our models.  In addition, I explore how statistical temporal 
properties impact routing protocols.  I analyze one-to-one routing 
schemes and develop new routing algorithms that consider 
autocorrelation, and reverse link and consecutive same path link lagged 
correlations.  I have developed two new routing algorithms for the cost 
link model: (i) a generalized Dijkstra algorithm with centralized 
execution, and (ii) a localized distributed probabilistic algorithm.

Finally, I describe all the insight obtained by analyzing all these 
data, and propose new research direction for network protocol and system 
designers of sensor networks.

Short bio:
Alberto Cerpa is a Ph.D. candidate in Computer Science at the University 
of California, Los Angeles (UCLA), working under the supervision of 
Professor Deborah Estrin.  Since Fall of 2000 he has been working in the 
Center for Embedded Networked Sensing (CENS) and the Laboratory for 
Embedded Collaborative Systems (LECS) at UCLA. Cerpa received his M.Sc. 
in Computer Science from the University of Southern California (USC) in 
2000, his M.Sc. in Electrical Engineering from the University of 
Southern California (USC) in 1998, and his Engineer Degree in Electrical 
Engineer from the Buenos Aires Institute of Technology, Argentina in 1995.

Alberto Cerpa interests lie broadly in the computer networking and 
distributed systems areas.  His recent focus has been systems research 
in wireless sensor networks, with emphasis in wireless radio channel 
measurement and characterization, network self-configuration, topology 
control, routing algorithms, programming models, and development of 
wireless testbeds.  He is also interested in Internet protocols and 
operating systems issues.  In the past, he has been involved in active 
networking, mobile IP, and protocol design and verification research.