Page 326 - Mechatronics for Safety, Security and Dependability in a New Era

Ch63-I044963.fm Page 310 Thursday, July 27, 2006 12:08 PM
Ch63-I044963.fm
310
310 Page 310 Thursday, July 27, 2006 12:08 PM
movements. Communication with RFID (Radio frequency identification) is not bi-directional (for
example, see Hightower et. al. (2000)), and sensitive range of passive RFID is too short for tracking
people. Wireless LAN devices consume higher electrical power (typically about 1[W]) than devices of
other methods. Moreover, TDOA (Time difference of arrival) system becomes expensive because it
needs high-resolution clock to measure small arrival time difference of radio wave. This paper
describes development of a positioning system based on measurement of propagation loss of Bluetooth
wireless transmission.

POSTPONING SYSTEM USING BLUETOOTH

In our proposed system, several base stations with Bluetooth wireless module are placed in the
measurement area, and a person carries a mobile unit also equipped with a Bluetooth module. Pseudo
distances between the mobile unit and the base stations are derived from the propagation losses.
Position of the mobile unit is calculated from these pseudo distances using a trilateration algorithm
(described in Kitasuka (2003)).
The relationship between the received signal power Pr[W] and the transmission signal power Pt[W] in
free space at distance d is calculated from a propagation model. We used the free space model and the
two-ray ground propagation model that account for both the direct line-of-sight path and the path
reflected from the ground (described in Fall (2001)). Since real propagation is a combination of the two
models, we will introduce constants a and ft to combine the two models to express a practical
propagation model defined by the following equation.

'-$
Constant a depends on the system, such as gains of antennas and the wavelength, and constant fi
depends on the environment. The experimental data have shown good agreements with the proposed
propagation model. The values of the parameters obtained by curve fitting on the experimental result,
were a = 2.01 and /5=2.59. Theoretically, the propagation models mentioned above are only
applicable to far filed region. In the present case, far field region begins approximately 1.2[m] from the
antenna, which is 10-wave length of 2.45[GHz] radio wave.

Influence of the human body and variance of propagation loss

Since the frequency of Bluetooth is 2.45GHz, absorption by human body affects the distance
measurement based on propagation loss. Figure 1 (a) shows the relationship between the distance and
the propagation loss when a human body is in the propagation path. A person with a mobile unit on the
front of the body stood away from a base station, and propagation loss was measured as the person
turned 90, 135, and 180 degrees, at distances of l[m] to 7[m] at l[m] interval. The angle facing the
base station was defined as 0°. When the angle was larger than 90°, the mobile unit lost the line of
sight to the base station and the propagation loss increased by more than 10 [dB], which was equivalent
to an error of approximately 10% of the pseudo-distance. This result indicates that compensation for
the human body influence is necessary.
Figure 1 (b) shows the variance of propagation loss at distance of 3[m] and 7[m], The result indicates
that the variance of propagation loss is smaller at shorter distance.