|
|
|
|
|
Our research spans a broad set of disciplines, but is generally
in the areas of communications and networking, and more specifically wireless.
All of this is really at the system level, where we mix theoretical design,
simulation-based evaluation and hardware experimentation, driven by a wide
range of applications.
|
Underwater Networking
This is a collaborative project with Scripps Institution
of Oceanography, the MAE dept, the CSE dept and Calit2. The goal is to develop
networked swarms of underwater sensor drogues that drift freely along with
the currents. Specifically, we focus on the communication, localization and
networking aspects.
|
Experimental Platforms
Experimental validation is an important element in
systems research. We have started several prototype development efforts and
experimental studies. For example, at Calit2, we have developed CalRadio,
which is an 802.11 platform with a software programmable MAC layer.
|
|
Sensor Network Sampling
An important class of sensor networks aims at estimating
the spatiotemporal behavior of a physical phenomenon, thereby acting as
a distributed sampling system. We leverage the inherent spatial redundancy
to enhance the operational lifetime of the system.
|
Ad-Hoc Networking
We envision more and more computing to be embedded
into our surroundings, which can be accessed through wireless communications.
Our research addresses the resource allocation and scheduling challenges
that arise. Also, we are working on ad-hoc networking support for wearable
battefield sensing.
|
|
Mobile Sensor Networking
We investigate networking a small number of sensor-equipped
unguided aerial vehicles, and study related MAC, routing and scheduling issues.
Prototype systems based on small terrestrial robots are used as an experimental
platform.
|
|
|
Sensor Network Reliability
Due to their aggressive sizing and challenging operating
conditions, sensor networks are highly vulnerable to errors, both within
the nodes' hardware as well as the wireless links. We investigate the use
of data properties (i.e. correlation in sensor data) to achieve reliable
data collection within incurring additonal overhead on the nodes.
|
Sensor Network Wakeup Scheduling
In wireless sensor networks, energy efficiency is
crucial to achieving satisfactory network lifetime. To reduce the energy
consumption significantly, a node should turn off its radio most of the time,
except when it has to participate in data forwarding. Our technique, called
STEM, efficiently wakes up nodes from a deep sleep state without the need
for an ultra low-power radio.
|
Some more projects we worked on in
the past
- Energy Efficient Wireless Communications
- Sensor Networking: routing, addressing
- Turbo Coding
back to main projects
|
|
|
|
|
|
