Database Reference
In-Depth Information
3.5 Energy E
cient Design
Since sensors often have limited battery power, it is critical to design
the participatory sensing systems in order to maximize the life of the
system. Many critical sensing components such as GPS tend to consume
a lot of power, and can therefore result in a short life-cycle for the
system. A number of tricks can be used in order to reduce the energy
consumption, which may involve a reduction in the sampling rate at
which the data is collected. This reduction in the sampling rate can
often be achieved with the use of side-information which is collected
through more ecient means. Some examples of common tricks which
are used in order to improve the energy eciency are as follows:
In the
Sensloc
system proposed in [97], a GPS device, an accelerom-
eter, and a WiFi scanner are simultaneously used in order to de-
tect particular variables such as location with good accuracy. We
note that the energy requirement of different kinds of sensors may
vary, and the accuracy of the different sensors may also vary in
a time dependent way. The typical tradeoff is that while GPS is
extremely power hungry, it is also more accurate. Therefore, at a
given time, the data is selectively sampled at varying rates from
different sensors in order to fuse the measurements together, and
provide an accurate estimation of the desired variable without too
much energy consumption.
A rate-adaptive positioning system known as
RAPS
is proposed
in [129]. The system is based on the observation that GPS is gen-
erally less accurate in urban areas, and therefore it makes sense
to turn it on only as often as necessary to achieve this accuracy.
The
RAPS
system uses the location-time history of the user to
estimate user velocity and adaptively turn on GPS only if the es-
timated uncertainty in position exceeds the accuracy threshold. It
also estimates user movement using a duty-cycled accelerometer,
and utilizes Bluetooth communication to reduce position uncer-
tainty among neighboring devices. In addition, the system em-
ploys celltower-RSS blacklisting to detect GPS unavailability, in
which case it is not turned on at all. Thus, the use of context-
sensitive information in order to adaptively turn on GPS results
in a considerable about of power savings [129].
Often, a large amount of rich context sensitive information is avail-
able, which can be used to improve the accuracy of the sensing
measurements without spending additional energy. For example,
if the last known GPS location is overlaid on a map, then the fu-
