| |
 |
The Protean* Living Project is exploring technologies
to enable a network of distributed devices to seamlessly cooperate to
simplify daily tasks.
|
News
- [10.09.06] New meeting time: Mondays 1-2:30pm, BWRC Video Conference Room.
- [10.05.06] Updated research abstracts
posted.
- [09.05.06] New meeting time: Fridays 2:30-4pm,
BWRC round table.
- [06.28.06] Agreed upon project code-name,
initial web-page deployed, new material posted, and under continued
construction.
Faculty
| Jan M. Rabaey |
Professor, UCB |
| John Wawrzynek |
Professor, UCB |
| Adam Wolisz |
Adjunct Professor, UCB; Professor TU Berlin |
| Jan Newmarch |
Visiting Professor, Monash University |
Graduate Students
| Chris R. Baker |
crbaker@eecs |
| Yury Markovsky |
yurym@eecs |
| Jana Van Gruenen |
janavg@eecs |
| Stanley Chen |
yschen@eecs |
|
@eecs = @eecs.berkeley.edu
|
|
Undergraduate Students
| Donald Szeto |
| Bobak Mortazavi |
The Protean Living Project is an umbrella for several research angles:
system integration, transcoder development, system architecture,
programming abstractions, sensor networks, and novel hardware platforms.
Home Gateway Demo
|
This is an ongoing project to develop an ad-hoc distributed media
networking framework. This framework consists of media sources and
sinks, media streaming and transcoding, and the necessary mechanisms
for control. Some of these elements are implemented on a central
computational hub (or gateway), while others are distributed throughout
the network. Some of the key challenges are building a robust and
scalable infrastructure (i.e. its desirable to be able to "drop" new
modules into this framework), use of FPGAs in the compute hub and their
dynamic reconfigurability (to meet transcoding needs), and efficient
control over media streams.
|
|
|
Contact: Chris Baker
|
A System Architecture for Ambient Intelligent Environments
|
Although electronic devices permeate our surroundings and offer
unsurpassed power and features, many impediments still exist to realize
the concept of ambient intelligent environments (e.g. smart-buildings):
configuration complexity, minimal device interoperability, difficulty of
use, lack of personalization, limited automatic adaptation, security
and encryption, and little integration with sensors for awareness,
are only some of the problems faced. We define four tenets of ambient
intelligent environments: Zero-configuration, Universality, Multi-user
optimality, and Adaptability (ZUMA), and describe a system architecture
which embraces these concepts. We have developed a set of clean
abstractions for users, content, and devices. Furthermore, we have
developed a platform which enables configuration and organization of
content and networked heterogeneous devices in an ambient intelligent
environment. Finally, we have validated several aspects of this platform
with a prototype implementation and are working to analyze its flexibility
and infrastructure requirements.
|
|
|
|
Contact: Chris Baker
|
High-level Abstraction for Sensor Networks
The sensor network platform (SNSP) has a service-oriented
architecture; the capabilities of the sensor network are abstracted
as services. Services may also generate content that can be referred
to later (for example historical temperature readings). Another key
component of the SNSP is the inclusion of personae. In any application
space, there will be users that interact with the sensor network. The
concept of personae captures users’ preferences and their access to
certain functionality.
- Fundamental value (e.g. temperature or light intensity)
- Security and access restrictions
- Service invocation interface
- Structure of the service
- Calibration and data pre-processing
- Performance metrics
At the heart of the SNSP is a repository service that maintains copies
of all the services in the network. Currently, effort is being spent
investigating lightweight data replication techniques to store this
repository in the sensor network. The SNSP also allows bridging to
across node platforms, and more specifically integration with multi-media
applications.
|
|
|
Contact: Jana van Gruenen
|
Low Complexity Implementation of Video Transcoding with Access Control
|
As the number of networks, types of devices, and video coding standards
increase, interoperability between different video systems and
different networks is becoming more important. Video transcoding
is the key technologies that enable a seamless interaction between
different systems. The target of this research is to develop an efficient
real-time video transcoder for Home Gateway Demo. Intra-Refresh
transcoding architecture with selective encryption scheme is proposed
as an efficient solution for low complexity implementation. Dynamic
bit-rate and resolution adaptation provide transmission bandwidth saving
for various video clients with different processing power and display
capabilities. The integrated selective encryption scheme protects the
transcoded streams from unauthorized access. This project will use FPGA
platform for implementing the prototype system and demonstrating the
design concepts.
|
|
|
Contact: Stanley Chen
|
Universal Content Router
Technology advances are beginning to make ubiquitous computing a
reality. The devices in this environment will be heterogeneous in
many regards; we are concerned with their means of communication.
To enable interoperability among these devices, all levels of the
networking stack require translation. Several solutions can address this
interoperability challenge:
-
A stovepipe solution: A single company or group of companies
select a set of vertically integrated standards. Compliant devices
are ensured interoperability on all levels. The problem with this
approach is that they are exclusive in the covered functionality and
do not allow for dynamic expansion of this functionality. Furthermore,
the diverse needs of different devices in terms of bandwidth, latency,
power consumption, and cost preclude the success of this approach.
-
The flexible peripheral solution: Peripheral devices can be made
flexible and adaptable so that they can support a variety of protocols
and formats (\emph{e.g.} multi-standard cell phones). The disadvantage
of this approach is cost, power consumption, and inability to upgrade.
-
The network overlay solution: This approach uses bridges to
connect different networks. This solution removes the burden from the
end devices, placing it on the infrastructure.
The approach advocated in this research presents a powerful
generalization of the third option by concentrating processing
power in a central unit, which we term a Universal Content Router,
or UCR. The UCR would be responsible for enabling interoperability
between devices for all layers of the networking stack, from physical
layer (e.g. with software-defined radios) to application layer.
From a system management point of view a central unit is simpler to
deploy, update, and maintain, than a heterogeneous collection of devices.
The economics of centralization are also favorable; after the initial
cost of infrastructure deployment, peripheral units can be added at low
cost. Furthermore, this central unit is extensible (both hardware and
software) so that it can accommodate future standards and peripherals.
This research is focused on the hardware architecture of a UCR and how
to manage its resources in software.
|
|
|
Contact: Chris Baker
|
2006
|
Chris R. Baker, Yury Markovsky, Jana Van Gruenen, Adam Wolisz, Jan
Rabaey, and John Wawrzynek. ZUMA: A Platform for Smart-Home
Environments. In Proceedings of IET Intelligent
Environments. Athens, Greece. July, 2006.
|
 
|
|
Jana Van Gruenen, Yury Markovsky, Chris R. Baker, Adam Wolisz, Jan Rabaey,
and John Wawrzynek. ZUMA: A Platform for Smart-Home Environments,
The Case for Infrastructure. In Proceedings of IET Intelligent
Environments. Athens, Greece. July, 2006.
|
|
|
|
| Universal Content Router, BWRC Summer Retreat [06.05.06] |
| |
| Home Gateway Demo |
|
| Wireless Home Infrastructure, World Wireless Forum |
|
| GSRC Quarterly Symposium, Berkeley CA [03.14.06] |
 |
| AIE Related Work [06.28.06] |
| |
| Reading List |
 |
| Postal Mail: |
2108 Allston Way Suite 200 Berkeley, CA 94704 |
| E-Mail: |
crbaker@eecs |
Funding Sponsors
This project is supported by the Gigascale Systems Research Center (GSRC) and the Berkeley Wireless Research Center (BWRC).
* In Greek mythology, Proteus is an early sea-god, one of several deities
whom Homer calls the "Old Man of the Sea." He can foretell the future,
but will change his shape to avoid having to; he will answer only to
someone who is capable of capturing him. From this feature of Proteus
comes the adjective protean, with the general meaning
of "versatile", "mutable", "capable of assuming many forms," with
connotations of flexibility, versatility and adaptability.
| |
