image placeholder

Google’s mission is to “organize the world’s information and make it
universally accessible and useful.” To help succeed in this mission,
Google maintains strong ties with academic institutions that are doing
world-class research. As one component, Google has developed the
Research Awards program to provide funding support to researchers at
these institutions.

Recently, three groups in the Computer Science Department at UCSB were
successful in receiving Google Awards for their world-class research
and education projects. One of these projects focuses on wireless
networks and cognitive radio, the second projects aims to build an
engine for scalable web applications, and the third project addresses
the problem of energy-efficient computing.

The first project, lead by Prof. Elizabeth Belding, is titled
“Online Spectrum Auction Systems.” Although cognitive radios and
opportunistic unlicensed spectrum access have become popular research
topics of late, few, if any, of the proposed solutions have been
implemented and tested in practice. The time is right for the
evolution from analytical analysis and simulation to implementation to
occur, due to both the recent availability of suitable hardware
platforms, as well as the maturation of spectrum access and auction
algorithms. To this end, the researchers propose to build a prototype
of an operational online spectrum auction system, based on previously
published auction algorithms. In so doing, they will address many of
the implicit or explicit assumptions and simplifications made in the
original system design, and solve the real-world problems surrounding
the design and implementation of such a system, thereby demonstrating
its viability.

The second project is called “GRAEScale: The Google Research AppEngine
for Development and Deployment of More Diverse and Scalable Web
Applications.” It is lead by Prof. Chandra Krintz and Prof. Rich
Wolski. The Google AppEngine (GAE) has lead the way in simplifying the
development of web applications and exposing large-scale compute
resources to these applications transparently. The goal of the work in
this project is to increase the potential of such an approach. To this
end, the researchers propose GRAEScale, a set of novel extensions to
GAE that facilitate development of a greater diversity of scalable GAE
applications, and that provide a framework for researchers,
developers, and students to investigate new technologies and
techniques for scalable web services for cloud computing. As part of
the work, GRAEScale will be integrated with Eucalyptus, an extensible,
open-source cloud software architecture that was previously developed
by the group at UCSB. An initial prototype of this integration effort,
which is called AppScale, can already be checked out here.

The third project, lead by Prof. Fred Chong, is titled “An Innovative
Multi-disciplinary Graduate Course in Energy-Efficient Computing.” As
computing and data storage increasingly become a globally-available,
public utility, the proliferation of large numbers of servers and
massive data centers will have a substantial energy footprint in our
future. Interestingly, energy expenditures are already becoming more
significant than the cost of machines. This suggests that the US will
have a critical need for scientists trained to reason about
energy-efficient computing. To tackle this challenge, the proposed
project takes a first step by creating a graduate course in
energy-efficient computing. To this end, the newly-formed UCSB
Institute for Energy Efficiency will be leveraged, helping to bring
together knowledge from Computer Science, Electrical and Computer
Engineering, Mechanical Engineering, and the Bren School of
Environmental Science and Management.