Instrumentation & Measurement Magazine 26-4 - 23

Fig. 3. Schematic representation of the adopted set-up.
[29]. Fig. 3 shows a schematic
representation of the
set-up installed within a localization
domain having
dimensions 6 m x 6 m x 1.5
m. In this design, five UWB
devices are used: four devices
have been configured
as anchors (fixed inside the
localization domain), while
another device, whose position
must be estimated,
has been configured as a
tag (free to move within
the localization domain).
All anchors are powered
by portable batteries while
the tag is connected to a
raspberry pi 4 which is also
powered by portable battery.
The raspberry pi is
Ultra-Wide-Band-based Indoor Localization
In this section the use of Ultra-Wide-Band (UWB) wireless
communication technology for locating devices in indoor
environments is described. Localization represents the
process by which a device (tag) is able to dynamically orient
itself in a given environment. While the use of Global
Navigation Satellite Systems is well established in outdoor
scenarios, the research for methodologies able to track the
tag position efficiently in indoor environments is an open
topic. This is due to the greater complexity of indoor environments,
given the presence of obstacles, walls and
people in motion which cause interference phenomena, signal
attenuation and multipath propagation. Today, there
are many contexts that require the localization task to be
solved: some examples concern applications related to the
IoT paradigm, industrial automation, medical assistance
and security services.
The proposed indoor localization system consists of several
UWB devices. In particular, some are configured as anchors
and are fixed devices with known positions, and then there
is a tag whose unknown position must be estimated. The
TREK1000 kit from Decawave was used to implement the experimental
set-up. TREK1000 devices are compliant with the
IEEE 802.15.4-2011 UWB standard, support different data
rates (110 kbps, 850 kbps and 6.8 Mbps) and use six frequency
bands from 3.5 to 6.5 GHz. Through bi-directional anchor-tag
communication, and adopting two-way ranging time-of-flight
measurements, it is possible to estimate the anchor-tag distance.
Once all of the anchor-tag distances have been estimated, since
the anchors positions are known, the unknown position of the
tag is obtained by solving a numerical optimization problem
June 2023
used to collect all estimated anchor-tag distances as well as
compute the tag position in real time condition. In fact, the
entire localization algorithm is implemented on board the
raspberry pi. More details on the arrangement of the set-up are
shown in [30].
To verify the performance of the developed localization
system, two types of tests were carried out. A static test, in
which the tag remains stationary in a given position, and a
dynamic test, in which the tag follows a path, were considered.
Regarding the static test, ten test positions within the
localization domain were considered. For each of the test
positions, the performance of the localization system was
evaluated, and the obtained results showed a maximum
localization error, i.e., the maximum of errors on all test positions,
equal to 30 cm. Subsequently, the dynamic test was
performed in which the tag was moved along a rectangular
path, and four waypoints were used to evaluate the localization
error. In this case, the obtained maximum localization
Instrumentation & Measurement Magazine 26-4

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 26-4
