Active RFID Positioning of Vehicles in Road Traffic

Enzhan Zhang
School of Communication & Information Engineering
University of Electronic Science and Technology of China
(UESTC)
Chengdu, 611731, China
[email protected]
Weili Jiang
Communication and Information Engineering
Chongqing University of Posts and Telecommunications
(CQUPT)
Chongqing, 400065, China
Yujun Kuang, IEEE member
School of Communication & Information Engineering
University of Electronic Science and Technology of China
(UESTC)
Chengdu, 611731, China
[email protected]
Mifta Ahmed Umer
School of Communication & Information Engineering
University of Electronic Science and Technology of China
(UESTC)
Chengdu, 611731, China
[email protected]

AbstractVehicle positioning is critical for Inter-vehicle
Communication, navigation, vehicle monitoring and tracking,
which are regarded as the technology to make everyday-driving
safer. Conventional positioning methods (for example GPS)
have poor accuracy (approximately ten meters, D-GPS has
higher accuracy about one or two meters.), and could not
positioning onto the lane, which is not enough to satisfy vehicle
safety requirements such as collision avoidance.
RFID (Radio Frequency Identification) is an old but newly
emerged technology. In this paper, we propose RFID technology
into vehicle positioning to improve positioning performance. The
idea of this positioning scheme is that the RFID tags are installed
along lanes on road to form RFID matrix and the vehicle
equipped with RFID readers locate itself by retrieving position
information from tags around it, thus called Active RFID
Positioning (ARP).
The proposed ARP scheme not only dramatically improves the
accuracy of positioning; it is also more practical as well for
scenarios like tunnels or underground parking where GPS is not
available now.
Keywords- vehicle positioning; active RFID-based positioning;
RFID
I. INTRODUCTION
Future active vehicle-safety applications rely heavily on
precise real-time positioning. Many approaches have been
proposed for positioning vehicles, such as GPS or A-GPS
based positioning [1, 2] for vehicle navigation or emergence
applications, and for optimization purposes in mobile networks
[3] and Wireless Sensor Networks (WSN) [4, 5]. By using
GPS positioning [1, 2], vehicles can be located with error
approximately ten to twenty meters, or even better by
Differential GPS (D-GPS) with errors about one to two meters.
However, these GPS-based positioning approaches cannot
provide enough precision to lock vehicles on their lanes in
addition, with GPS it is difficult to get position in tunnel or in
complex downtown areas surrounded by skyscrapers, or along
multi-layered high-speed tracks. Recently, location based
services grow fast by using hot-spots such as WiFi access
networks, which provides much poorer precision than GPS
(typical error will be several hundreds meters) and RFID is
also used for positioning, which is common in Automatic Toll
Collection on highway networks however, it has lower
accuracy in compare with GPS, and it is usually available only
near the entries or exits of highway networks, which can be
viewed as passive positioning in viewpoint of vehicles, where
RFID readers are installed at Toll Stations and intersections etc.
so as to read the data stored in the tag attached on the front
windshield of vehicles. But such systems cannot provide any
position information when the vehicle is outside the range of
readers.
Vision based positioning is an active positioning method [6,
7]. In this method, a single camera is used to determine
vehicles lateral and longitudinal offsets with respect to the
road. Line features on the road are detected to estimate
vehicles global pose. Obviously, this method can not give
global coordinates and can not work without road marks.
Papers [8-10] proposed an active RFID location system for
indoor/in-building environments. The paper [10] introduced an
idea of using RFID technology for positioning on road and
presented preliminary results on RFID communication speed
and tag database access time.
In this paper, we proposed Active RFID Positioning (ARP)
scheme where RFID tags are deployed along each lane of the
road. However, we will focus more on tag and reader
installation, the positioning error performance. Results show
that, the new positioning method ARP is practicable to provide
enough vehicle positioning precision for vehicles to determine
their position in which road segment and on which lane.
The rest of this paper is organized as follows. In Section II,
the scheme of ARP is given. Section III introduces the
installation and arrangement of tags. The performance of the
method is analyzed in Section IV. Finally, conclusions and
future works are described in section V.
This work is supported by National Natural Science Foundation of China
(61071099)
The 11th International Symposium on Communications & Information Technologies (ISCIT 2011)
978-1-4577-1295-1/11/$26.002011IEEE 222
II. ACTIVE RFID POSITIONING SCHEME
Fig. 1 gives a systematic view of the proposed ARP
scheme, and a typical arrangement of RFID tags along one
direction of the road. The vehicles mounted with RFID tag
readers continuously generates RF carrier waves while moving
along the road, thus information contained in the RFID tags
(small rectangles on lanes) will be retrieved. By collecting data
from tags around the vehicle it can estimate its position and
speed, which is fed into the GIS based navigation system and
exchanged with other vehicles nearby by wireless
communication. Then with vehicle status and that of its
neighbor, it can evaluate the traffic situation nearby, forecast
vehicle collision or dangerous driving pattern and
automatically update optimal routes or activate alarms when
necessary. Two important aspects of the proposed ARP are
installation/arrangement of RFID tags along the road and
characteristics requirement of tags and readers. Although
passive (without battery) tags are quite cheap and pervasive
applications, large number of tags is required, so that tag
matrix pattern must be carefully designed for reasonable
position precision in order to minimize the overall investment
(We will study it in our future work.). Besides careful budget
planning of tag arrangement, this paper mainly focuses on tag
and reader installation, positioning precision, and evaluating
the relationship among reading range, vehicle speed, tag
installation and inter-tag distance.

Figure 1. Active RFID positioning
In the following, we assume that RFID tags contain data
including its position belonging lane and road, which can be
used to match positions in geographic information system
(GIS), thus vehicles can determine their positions.
III. INSTALLATION AND ARRANGEMENT
In this section, we present detailed installation of tag and
reader, and give basic patterns of tag arrangement.
A. Tag and Reader Installation
It is obvious that the readers should face down to the tags
deployed along the road, but if the orientation is oblique to the
road, it has great impact on the reliability of reading data
because of the vehicle movement and limited speed of data
retrieval. In the following, as shown in Fig. 2, we consider
three cases of tag and reader installation. In this paper, the
reader is installed on the chassis under the vehicle or on the
bumper at the front of the vehicle.
1) Facedown installation
In this case, the main lobe of tags is facing upwards while
the readers main lobe faces downwards as shown in Fig. 2-(1).
In Fig. 3, we assume that the vehicle is moving from left to
right with speed v, the readers main lobe width is o, and the
tags is |, the readers height is H from the ground. When the
tag is covered by the readers beam (In this paper, we call that
the reader sees the tag.), and the reader enters the coverage
of the tags beam (We call that the tag sees the reader.) in
the same time. This means that they are contacting with each
other. L is the distance from the entry of the read range to the
exit of the read range. Here, read range is the range in which
the tag and the reader can see each other. And we define
contacting time t
c
is the time in which the reader and the tag
can see each other. Thus from Fig. 2-(1) we obtain the
maximum contacting time as:

( )
( )
tan 2 2 ,
2 tan 2 ,
c
v
v
L H
H
v
t
| o |
o o |
>
= =

<

(1)
o
|
|
|
o
o
u

Figure 2. Tag and reader installation
2) Incline Forward
In Fig. 2-(2), the readers main lobe faces downward and
forward to tags ahead. We have two cases as follow:
a) Case 1
| o >

In this case, o is bigger than |. So, the reader always
sees the tag first. And then the tag sees the reader when
the reader enters the tag beams coverage. At the end of the
contacting, the reader can not see the tag last.
In Fig. 3, D is the distance between two adjacent tags along
the same lane. In fact, the readers electromagnetic wave
propagation range is larger than the tags. So, R is the reading
range (or reading distance,), which is determined by the
propagation range of tags electromagnetic wave.u is an angle
between the readers main lobe direction and the vertical. is
an angle between the tags main lobe direction and the road
surface. Where, 0<u<t/2, 0< <t/2, and u+=t/2.
223
When ( ) 2 arcsin H R | > as shown in Fig. 3-(1), is
big in this situation. So, the reader can only enter the tags
coverage from flank. Base on the above assumption, the
maximum contacting time can be obtained as:

( ) ( ) 1 tan 2 1 tan 2
L
t
c
v
H
v
| | +
=
= (

(2)
When -|/2<arcsin(H/R) as shown in Fig. 3-(2), is small
in this situation. So the reader enters the tags coverage from
front. The maximum contacting time can be obtained as:

( )
1
2 2
tan 2
L
t
c
v
R H
v
H |
=
=
(
+
(

(3)
( ) R H arcsin 2 > |
( ) R H arcsin 2 < |
u
o

|
|

u
o

Figure 3. case 1
o | >

b) Case 2
| o <

In this case, who sees the other first, the reader or the tag,
that depends on angle .
When ( ) R H arcsin 2 > | as shown in Fig. 4-(1), it is
similar to the situation which is shown in Fig. 3-(1). The reader
enters the tags coverage from flank. In this scenario, the tag
sees the reader first, and later, the reader sees the tag. At
the end of the contacting, the tag is out the readers sight
first. With the above assumption, the maximum contacting
time can be obtained as:

( ) ( ) 2 2 tan tan
L
t
c
v
H
v
o o u u
(

=
= +
(4)
When ( ) R H arcsin 2 < | as shown in Fig. 4-(2), we
have two scenarios.
(a) t/2+o/2-arccos(H/R)<<|/2+arcsin(H/R), as shown in
Fig. 4-(2)-(a), in this scenario the tag sees the reader first.
And then the reader sees the tag when the tag is covered by
the readers coverage later. On the contrary, the tag is out the
readers sight first.
(b) 0<<t/2+o/2-arccos(H/R), as shown in Fig. 4-(2)-(b),
in this scenario, the reader sees the tag at location 1 until the
reader arriving at location 3. And the tag sees the reader at
location 2 until the reader move out of the tags coverage. The
maximum contacting time can be obtained as:

( ) ( ) ( )
( )
tan 2 tan 2 , cos 2
1
2 2
tan 2 ,
L
t
c
v
H
R H
v
R H H else
v
u o u o u o
u o
=

+ > + (

=

(

(

(5)
( ) 2 arcsin H R | >
( ) 2 arcsin H R | <
cos( 2) R H u o < +
cos( 2) R H u o > +
u
o

|
|

o
u
|

o
u

Figure 4. case 2 o<|
3) Incline Backward
As shown in Fig. 2-(3), this case is just a horizontally
flipped version of which indicated in Fig. 2-(2), thus the
maximum contacting time is the same as in subsection 2).
B. Tag arrangement
According to different practical requirements, there are
many tag arranging methods. Hence, the following three main
factors should be considered in the tag arrangement.
1) Interference among adjacent lane tags must be
minimized;
224
2) Though the passive tags are cheap, large quantity of
tags are required in our system, thus tag density along the
road should be balanced as per cost and precision;
3) Distance among adjacent tags should match with the
contacting time and vehicle speed for optimal position
information sampling. See Fig. 5 where a typical arrangement
of tags is shown.

Figure 5. Tag arrangement
Base on this arrangement, we focus on positioning error
analysis of different installation methods of reader and tag in
the following section.
IV. ERROR ANALYSIS
In our system, the position of the vehicle is the information
which is read from the present tag, and the position
information will not to be updated until the reader has read the
next tag. The accuracy is determined by several key factors
(e.g. tags installation, arrangement, and reading range). In this
section, let the reading range be R. And D denotes the distance
between two adjacent tags along the same lane as shown in
above figures. We analyzed the positioning accuracy in
subsection IV-A, and investigated feasibility in subsection IV-
B, which is based on installations described above. The
comparison of ARP, GPS and wireless network positioning
(WNP) was given in subsection IV-C.
A. Accuracy analysis
1) Facedown
As shown in Fig. 2-(1), when the readers beam and the
tags beam are overlapping, that means the reader starts
reading the tags, and does not read two tags at the same time.
In this paper, we suppose the reader can obtain tags data when
passing by tag above. When the vehicle speed is v, the reader
will cost time T from starting to communicate with the tag to
getting the tags data. It can be seen from Fig. 2-(1), the
vehicles position has the maximum error at the moment
before the reader getting into the next tags beam coverage.
The maximum positioning error is given by:

( )
( )
0
0
0
max
tan 2 ,
tan 2 ,
2
T
T
T
D H
D H
D vdt
vdt
vdt
E L
| o |
o o |
+
+
= +
>
=
<

}
}
}
(6)
And, the minimum error is 0, when the vehicle is just
above the tag.
2) Inclined
a) Case 1
| o >

When ( ) 2 arcsin H R | > as shown in Fig. 3-(1), the
achieved maximum error is given by:

( )
( )
1 1
0
0
,
tan 2
tan 2
,
max
T
case
T
H
H
D vdt D d
E
vdt else
|
|

+ >
=
+
}
}
(7)
where, ( ) 1 1
0
2 tan 2
T
case
d H vdt |

(
=
(

}

When ( ) R H arcsin 2 < | as shown in Fig. 3-(2), the
achieved maximum error is given by:

2 2
1 2
0
2 2
0
,
,
max
T
case
T
D R H vdt D d
R H vdt else
E

+ >

}
}
(8)
where
( )
2 2
1 2
0
2
T
case
d R H vdt

=
}

b) Case 2:
| o <

When ( ) R H arcsin 2 > | as shown in Fig. 4-(1), the
achieved maximum error is given by:
( )
( )
2 1
0
0
2 ,
2 ,
tan
tan
max
T
case
T
D H vdt D d
H vdt else
E
o
o
u
u

+ + >
+
}
}
=
(9)
where ( )
2 1
0
2 tan 2
T
case
d H vdt u o

= +
(

}

When ( ) R H arcsin 2 < | , which is shown in Fig. 4-
(2). If ( ) cos 2 H R u o + s , as shown in Fig. 4-(2)-(a), the
achieved maximum error is given by:
( )
( )
2 2( )
0
0
,
,
tan 2
max
tan 2
T
case a
T
D H vdt D d
E
H vdt else
u o
u o

+ + >
+
}
}
=
(10)
where ( )
0
2 2( )
tan 2 2
T
case a
H d vdt u o

+
(
=

}
.
If ( ) cos 2 H R u o + > , as shown in Fig. 4-(2)-(b), the
achieved maximum error is given by:

2 2
2 2( )
0
2 2
0
,
,
max
T
case b
T
D R H vdt D d
R H vdt else
E

+ >

}
}
=
(11)
where
( )
2 2
2 2( )
0
2
T
case b
d R H vdt

=
}
.
B. Feasibility analysis
The reading range of passive RFID tags is about six meters
(The range depends on environment, tag type, etc.). In this
subsection we analyzed the relationship between the
contacting time and the vehicle speed, which is based upon
225
the following assumptions: (1) reading range R=6m; (2) reader
height H=0.15m; (3) tag and reader beam width =60
o
; (4)
data retrieval time (or a reading transaction time) is 20ms (a
little bigger than experiment result 18ms in paper [10]).
1) Facedown case
0 5 10 15 20 25 30
0
10
20
30
40
50
60
70
80
90
100
Vehiclespeed(m/s)
c
o
n
t
a
c
t
i
n
g

t
i
m
e

(
m
s
)

Figure 6. Relationship between contacting time and vehicle speed
According to the geometric relationship in Fig. 2-(1), we
calculated the longest distance L=0.173m, when the reader
passes through the tags beam. Fig. 6 illustrates the
relationship between contacting time and the vehicle speed.
As can be seen from Fig.6, the reader can not get tags data
when the vehicle speed is higher than 8.5m/s.
In order to provide enough time for contacting, we gave
incline forward installation to increase the distance L in
subsection III-2) and III-3).
2) Inclined case
0 10 20 30 40 50 60 70 80 90 100
0
100
200
300
400
500
600
700
800
900
1000
Vehiclespeed(m/s)
c
o
n
t
a
c
t
i
n
g

t
i
m
e

(
m
s
)

Figure 7. Relationship between contacting time and vehicle speed
According to the geometric relationship in Fig. 2-(2), we
calculated the distance L=D-d=5.91m (approximately six
meters). Under the same assumptions, the relationship between
contacting time and vehicle speed is shown in Fig. 7. The
result illustrates that the reader can get all tags data when the
vehicle speed is lower than 295.5m/s. Thus if the vehicle speed
is lower than 60m/s that the reader can have more than 98.5ms
to spend on reading tag data. So, the proposed positioning
scheme ARP is feasible.
C. Comparison
We give a comparison of ARP, GPS and wireless
network positioning (WNP) in table.1, which bases on papers
[11-24].
Table 1. Comparison of ARP, GPS and WNP
ARP GPS WNP
Infrastructure
RFID
tags
satellites
Wireless
networks
Map yes yes yes
Accuracy (meters) <6m >1m >35m
Positioning to lane yes no no
Time-consuming
short
(<2s)
long
(<5s)
medium
(<3s)
Anti-destroying ability strong medium weak
Sky blocked scenario available disabled available
Coverage narrow wide medium
Network dependence no no yes
Cost cheap expensive medium
Complexity low high medium
V. CONCLUSION AND FUTURE WORK
In this paper, a new positioning scheme ARP was proposed.
We first show all kinds of installation methods of tag and
reader. And then, the positioning error was analyzed under
different installation methods. At last, the comparison of ARP,
GPS and WNP was given. The results and the comparison
show that the method is feasible.
The proposed ARP method can provide information about
vehicles location on which lane and in which road segment.
Therefore, the method can be used in VANET for vehicles
lateral offset warning; lane based or road segment based
emergency warning messages (EWM) broadcasting; and
position based nodes clustering. For example, the reader can
periodically scan the tags when the vehicle moving above tags
along the road, and obtain its location, which is feed to the
vehicular CPU system. If the system does not receive the same
lane information during preset duration, it will assume that the
vehicle deviates from previous lane and can activate alarms by
voice or indicator light to warn the driver timely. Or if the
reader suddenly gets another other lanes information, the
alarm provides Lane changing warning to the driver. In
highway traffic, the warning can reduce traffic accidents,
which is caused by sudden lane changing due to fatigue
driving etc.
Although the ARP scheme is feasible, there are still many
issues to be resolved before practical application such as exact
cost control, and tag coding which will be investigated in the
future.
226
REFERENCES
[1] Capkun S, HamdiM, Hubaux J P, GPS-free positioning in mobile ad-
hoc networks, Cluster Computing, Vol.5, No.2, pp.157-167, 2002.
[2] Gui Y J, Ge S S, Autonomous vehicle Positioning with GPS in urban
canyon environments, Proceedings of the 2001 IEEE International
Conference on Robotics & Automation. Seoul: Korea, pp.1105-1110,
2001.
[3] Shi Sheng-li, Xu Jian-min, Qin Zhong, Vehicle Positioning Based on
CDMA and GSM Networks, Journal of South China University of
Technology (Natural Science Edition). Vol. 35, No. 2, pp.50-53, Feb.,
2007.
[4] Ni Hai-yan, Ying Xiang-yue, Jian Jia-wen, Hu Jun-ming, Study on
Self-positioning Algorithm for Wireless Sensor Networks, JOURNAL
OF NINGBO UNIVERSITY(NSEE), Vol. 22, No. 1, pp.7-11, Mar.,
2009.
[5] Zhang Jia, Wu Yan-hai, Shi Feng, localization algorithm based on DV-
HOP for wireless for wireless sensor networks, Journal of Computer
Applications, Vol. 30, No. 2, pp.323-326, Feb., 2010.
[6] Bao Rong Chang, Hsiu Fen Tsai, Chung-Ping Young, A Fast and
Effective Approach to Lane Marking and Neighboring Vehicles
Detections Based on Vision/GPS Sensing Together with Vehicle-to-
Vehicle Communication, 2009 Ninth International Conference on
Hybrid Intelligent Systems, pp.3-8, 2009.
[7] T. K. Xia, M. Yang, R. Q. Yang, Vision Based Global Localization for
Intelligent Vehicles, Intelligent vehicles symposium 2006, June 13-15,
2006, Tokyo, Japan, pp.571-576.
[8] K. Wendlandt, P. Robertson, M. Khider, M. Angermann, and K.
Sukchaya, Demonstration of a Real time Active-Tag RFID, Java Based
Indoor Localization System using Particle Filtering, in UbiComp,
Innsbruck, Austria, September 2007.
[9] Y. Huang, P.V. Brennan, A. Seeds, Active RFID location system based
on time -difference measurement using a linear FM chirp tag signal, in
Proc of PIMRC, 2008, pp.1-5.
[10] Hae Don Chon, Sibum Jun, Heejae Jung, Sang Won An, Using RFID
for Accurate Positioning, Journal of Global Positioning System Vol.3,
No.1-2, pp.32-39, 2004.
[11] Hui Hu, Chao Yuan, Performance analysis of Galileo global position
system, Power Electronics and Intelligent Transportation System
(PEITS), 2009 2nd International Conference on , vol.1, no., pp.396-399,
19-20 Dec. 2009.
[12] Chadwick D.J., Applications of the global positioning system to
intelligent vehicle highway systems, Telesystems Conference, 1993.
Commercial Applications and Dual-Use Technology, Conference
Proceedings., National , vol., no., pp.139-141, 16-17 Jun 1993.
[13] Porcino D., Location of third generation mobile devices: a comparison
between terrestrial and satellite positioning systems, Vehicular
Technology Conference, 2001. VTC 2001 Spring. IEEE VTS 53rd,
vol.4, pp.2970-2974 vol.4, 2001
[14] Calabrese F., Colonna M., Lovisolo P., Parata D., Ratti C., Real-Time
Urban Monitoring Using Cell Phones: A Case Study in Rome,
Intelligent Transportation Systems, IEEE Transactions on , vol.12, no.1,
pp.141-151, March 2011.
[15] Yang T.T., Kai Yan Yip, Use of MOBITEX wireless wide area
networks as a solution to land-based positioning and navigation,
Position Location and Navigation Symposium, 1994., IEEE, pp.91-98,
11-15 Apr 1994.
[16] Caffery J.J., Stuber G.L., Overview of radiolocation in CDMA cellular
systems, Communications Magazine, IEEE , vol.36, no.4, pp.38-45,
Apr 1998.
[17] Drane C., Macnaughtan M., Scott C., Positioning GSM telephones,
Communications Magazine, IEEE , vol.36, no.4, pp.46-54, 59, Apr
1998.
[18] Hung-Chi Chu, Rong-Hong Jan, Cell-based positioning method for
wireless networks, Parallel and Distributed Systems, 2002.
Proceedings. Ninth International Conference on, pp. 375- 380, 17-20
Dec. 2002.
[19] Gustafsson F., Gunnarsson F., Mobile positioning using wireless
networks: possibilities and fundamental limitations based on available
wireless network measurements, Signal Processing Magazine, IEEE,
vol.22, no.4, pp. 41- 53, July 2005.
[20] Chiao-Tzu Huang, Li-Wen Lo, Wei-ling Wang, Hsin-Lin Chen, A
study for optimizing the reading rate of RFID tagged cartons in
palletizing process, Industrial Engineering and Engineering
Management, 2008. IEEM 2008. IEEE International Conference on,
pp.1138-1142, 8-11 Dec. 2008.
[21] Bhadrachalam L., Chalasani S., Boppana R.V., Impact of RFID
technology on economic order quantity models, Systems Conference,
2009 3rd Annual IEEE, pp.327-332, 23-26 March 2009.
[22] Hori T., Wda T., Ota Y., Uchitomi N., Mutsuura K., Okada H., A
Multi-Sensing-Range Method for Position Estimation of Passive RFID
Tags, Networking and Communications, 2008. WIMOB '08. IEEE
International Conference on Wireless and Mobile Computing, pp.208-
213, 12-14 Oct. 2008.
[23] Donggeon Lee, Seongyun Kim, Howon Kim, Namje Park, Mobile
Platform for Networked RFID Applications, Information Technology:
New Generations (ITNG), 2010 Seventh International Conference on,
pp.625-630, 12-14 April 2010.
[24] Sharma M., Siddiqui A., RFID based mobiles: Next generation
applications, Information Management and Engineering (ICIME),
2010 The 2nd IEEE International Conference on, pp.523-526, 16-18
April 2010.
227