Navigation services which usually depend on GNSS are
limited to be used in open areas with satelite signals.If the
users or robots are about to move in buildings,another
approach must be used to navigate accurately. In our
approach Radio Frequency Identification (RFID) is used to
determine the location indoors.
In RFID positioning there are two common approaches to
estimate the location.One method is based on signal
strength.We take received signal strength indication(RSSI)
which present the power of received signal as the
measurement.Then the position is computed with certain
methods based on the measurements.Several methods have
been studied,such as RFID location fingerprinting,cell-based
positioning, and the way using ranges to the tags calculated
with RSSI.
Another particular method to estimate the location is
based on the landmarks.In the landmark-based
navigation,landmarks are required to be set in the
building,usually on certain doors and corners.A topological
map with nodes corresponding to the landmarks is used to do
the navigation.A navigation based on landmarks is without
physical reality.For example,a navigation may be go straight
till the end of the corridor,turn to the right,and stop after
passing two junctions.Thus it is important to make a suitible
topological map which should reflect the environment
perfectly.There must be enough number of landmarks to
complete the navigation without mistakes,so each landmark
should be unique in the map,easy to distinguish,and settled in
proper places such as doors or corners.With these demands
we propose an approach using radio frequency
identification(RFID) tags as landmarks.The RFID tags
communicate via electormagnetic waves,which means tags
can be read without contact and easily to distinguish from the
surroundings.The tags have unique identifications inside the
memories,and the memories may be read from several
meters away.