| |
Sensor Networks
Senor networks have quickly
become one of the most exciting and promising new technology areas.
If such networks are to be effective they must be:
 Maintenance-free
Deployed
easily
Simple
and redundant
Resilient
Resistant
to disruption & disturbance
JIGSAW makes possible the scenario where literally
tens of thousands of tiny, low power consumption, robust and
inexpensive sensors can be scattered across an area in any kind of
random distribution. Each sensor has no information about where it is;
it just sends a signal, which could be measuring local seismic
vibration, ambient noise, heat, moisture, movement etc. The receiving
system also has no information about the location of sensors, but uses
JIGSAW to build up a spatial mapping of the sensor zone on the basis of
correlation within the sensor signals. Natural forces and entropy may
destroy some sensors, but many more could be deployed to compensate. If
sensors are moved around, JIGSAW would simply keep updating the
sensors’ positions within its map.
<>JIGSAW enables a breakthrough in sensor networks
because it does not rely on known positions or special radio
intercommunications between sensors. This is a radical
departure from current technologies. However, it is also possible to
easily integrate JIGSAW into existing systems. For example, a standard
configuration of sensor/transmitters and one or more receivers could be
augmented with JIGSAW as a parallel system, able to jointly monitor the
incoming signals and derive an independent mapping of the sensors.
Differences between the assumed map and the JIGSAW mapping might
indicate faults in the signals or some kind of tampering. Such hybrid
systems are already being tested in our lab.
If sufficient constraint can be put on how the
system as a whole will operate it may be possible to dramatically
increase performance by optimising the algorithm within known
parameters.
|
|
