UCLA Internet Research Lab
GRAB Project
Objective
The objective of the GRAB project is to develop a set of sensornet protocols
that enables large-scale, configuration-free sensor deployment, reliable
sensor data collection in the presence of high levels of environmental noise,
and robust and efficient data reporting to highly versatile and mobile collection
points in spite of topological changes and component failures.
Characteristics of the Sensor Networks
The sensor networks for which the GRAB protocols are being developed can
be characterized in part by:
- A large number of sensors so as to prohibit the use of traditional
approaches for address allocation and routing.
- Sensors that have limited transmission range, and low peak
and
average data rates. (Due to power and size requirements.)
- A topology that constantly changes with the departure of depleted
sensors.
- A topology that may change with the addition of many new sensors.
- Sensors that are unreliable.
- Sensor stimuli that are unpredictable and mobile.
- Information demands that are unpredictable and mobile. (Unpredictable
in location and type of information.)
Design Goals
The design requirements for which the development of the GRAB protocols
will initially try to follow are:
- The protocol must scale well with the number of sensors.
- The resulting protocol overhead must fall within reasonable
bounds
dictated by the power requirements of a sensor network where
the
individual sensors will employ small batteries.
- The protocol should be robust against sensor failures and changing
environment conditions.
- The protocol should easily incorporate deployment of new sensors.
- The protocol should minimize delay in delivery of information
and provide
a reasonable probability that sensor information is actually
received by the end user.
- The protocol should minimize the amount of wasted effort due
to mobile
stimuli and mobile information sinks.
Two-tier Data Dissemination Model for Sensor Networks
Source code is available
here
. After downloading the .tgz file, put it in a directory, e.g.,
your home directory, then
tar xvfz ttdd-source-ns-allinone-2.1b8a.tgz
cd ttdd
./install
to compile the code.
Please read ~your-home/ttdd/ns-2.1b8a/padd/README for more information. Some
sample TTDD scripts are in ~your-home/ttdd/ns-2.1b8a/scripts. If you have
any questions or suggestions about the code, please send an email
to Fan Ye
.
Links