A Multi-Carrier UHF Passive RFID System

A Multi-Carrier UHF Passive RFID System
Hsin-Chin Liu, Yung-Ting Chen, and Wen-Shin Tzeng
Department of Electrical Engineering
National Taiwan University of Science and Technology
Taipei, 106, Taiwan
hcliu@mail.ntust.edu.tw
Abstract
The “Gen2” specification for UHF passive RFID
system released by EPCglobal has become an intense
research interest. A Gen2 tag derives its power from
the RF wave emitted by a Gen2 RFID reader and
responds its modulated backscatter signals to the
reader. Due to the large propagation loss, the
accessible range of a Gen2 tag is hence limited.
Moreover, the readability of a Gen2 tag is often
influenced by the multipath fading problem. In this
paper we propose a multi-carrier UHF passive RFID
system that can effectively extend the accessible range
of a Gen2 tag. In addition, we investigate the multiband
backscatter property of Gen2 tags, which can be
utilized to mitigate the multipath fading problems.
1. Introduction
The Radio frequency identification (RFID)
technology with its prosperous market and its diverse
applications has gained much attention recently. Owing
to the low cost, small size, long accessible range tags,
the UHF passive RFID system is especially promoted
by EPCglobal. In 2005, The “Gen2” specification for
UHF passive RFID system is released by EPCglobal
[1]. Because a Gen2 tag derives its power from the RF
waves emitted by a Gen2 RFID reader and responds its
modulated backscatter (MBS) signals to the reader, the
accessible range is hence limited due to the large
propagation loss. Besides the free space attenuation,
the multipath fading effect can often result in
readability problems.
In an ordinary EPCglobal Gen2 RFID system [1], an
integrated reader not only transmits continuous waves
(CW) to provide energy and the backscatter carrier to
tags, but also sends the pulse-interval encoding (PIE)
reader-to-tag (R-T) commands to tags and receives the
tag-to-reader (T-R) responses. Assuming that the
maximum accessible distance between a reader and a
tag is d
0
, and the power of the CW emitted from the
reader is P cw
as shown in Figure 1(a). In accordance
with the Friis equation [3], the received power of the
P can be presented as Eq.(1).
tag
tag
⎛ λ ⎞
P = P G G
tag cw ⎜ ⎟ , (1)
⎝4π
d
reader tag
0 ⎠
where λ is the wavelength of the CW, G reader
and
G
tag
are the power gains of the reader and tag
antennas respectively.
When the tag is powered up, it listens for the R-T
commands and adjusts its impedance match between
the IC and antenna appropriately to respond the MBS
T-R signals. The power of the MBS signal received at
the reader Prx
can be written as Eq.(2)
2
⎛ λ ⎞
Prx = Ptag ⎜ ⎟ GreaderG
. (2)
tag
⎝4π
d0
⎠
Substituting P from Eq.(1) into Eq.(2), we see
tag
that P
rx
is reverse proportional to the four squares of
d
0
.
The accessible distance of a Gen2 tag is mainly
determined by two factors: the power of the MBS
received at the reader P rx
and the received power of
the tag P tag
. For the former factor, a high sensitivity
reader can solve the weak P rx
problem. However, the
latter one is more troublesome. An intuitive solution of
increasing the P tag
without tag modification is to
enlarge the P cw
. However it is impractical due to
healthy concerns, the industrial, scientific and medical
(ISM) band compliance issues, and the restriction of
the local regulations.
In order to overcome the insufficient P tag
problem,
we propose a multi-carrier UHF passive RFID system,
2
Proceedings of the 2007 International Symposium
on Applications and the Internet Workshops (SAINTW'07)
0-7695-2757-4/07 $20.00 © 2007

whose diagrammatic sketch is shown in Figure 1(b).
The proposed system not only can provide sufficient
energy to Gen2 tags but also can extend its accessible
range. Furthermore, taking the advantages of multiple
carriers, the reader can be benefited from the frequency
diversity gain and the multipath fading problem can be
mitigated.
The remainder of this paper is organized as follows:
Section 2 introduces the proposed system. Section 3
presents the experimental results. Section 4 draws
conclusions.
2. The Multi-carrier UHF passive RFID
system
Unlike the integrated reader in an ordinary UHF
passive RFID system, the reader in a multi-carrier UHF
passive RFID system is decomposed into two parts: a
CW emitter (CWE) and a transceiver. The CWE can be
preset leaky cables or radio source(s) close to tags,
which emit a CW with frequency f
c
to illuminate tags.
The CWE constantly provides energy and the MBS
carrier(s) to the tags. The transceiver, which can be
farther away from the tags, is mainly used to transmit
the R-T commands and receive the T-R responses.
As illustrated in Figure 1(b), the received power of
the tag
P
'
tag
can be presented as Eq.(3).
2
' ' ⎛ λ ⎞
c
tag
=
cw ⎜ ⎟ cwe tag
4π
d1
P P G G , (3)
⎝ ⎠
where P denotes the power of the CW emitted from
'
cw
the CWE,
λ
c
is the wavelength of the CW, and d 1
is
the distance between the CWE and the tag.
G cwe
and
G
tag
are the power gains of the CWE and tag antennas
respectively. In addition, the power of the MBS
received at the reader
P can be written as Eq.(4).
'
rx
2
' '
⎛ λ ⎞
rx
=
tag ⎜ ⎟ rx tag
4π
d2
P P G G , (4)
⎝ ⎠
where G is the power gains of the transceiver, and
rx
d
2
is the maximum accessible distance between the
transceiver and the tag. Assuming that the antenna
gains of the CWE and the transceiver are the same as
that of the integrated reader in an ordinary system
( Gcwe = Grx = Greader
), and the CW frequencies and
powers in both systems are the same
'
( λ = λc,
Pcw = Pcw), the power of received MBS in
both system can be written as Eq.(5).
' 4
Prx
d0
= . (5)
2 2
Prx
d1d2
'
When d1 < d0
and Prx
= Prx
, d2 > d0
can be
derived. In other words, given the same receiver
sensitivity and the same CW emission power, the
proposed system can effectively extend the accessible
distance between a tag and the receiver. Moreover,
because the tags are plentifully illuminated by the
CWE, the problem of insufficient tag operation energy
is avoided.
A simplified normal operation flow of the proposed
system is illustrated in Figure 2. Before issuing an R-T
command, the transceiver turns on the CWE to emit the
CW, so that the tags nearby the CWE are powered up
and ready for the R-T command. When the accessing
operation is finished, the transceiver turns off the
CWE, and the operation is then terminated.
The proposed system may seem similar with an
ordinary Gen2 system, except the additional CWE. As
a matter of fact, they are substantially different. We
discuss three major differences as below:
2.1. The frequency difference between the
transceiver and the CWE
Because the transceiver and the CWE are
separated in our proposed system, the carrier frequency
of the transceiver ( f
t
) and the frequency of CW ( f
c
)
emitted by the CWE can be different. Assuming the
noise is negligible, the RF signal received by a tag can
be expressed as Eq.(6).
s t = m t cos 2π f t + Acos 2π f t+ φ t , (6)
() () ( t ) ( c ())
where
⎧A , when high level is sent
m()
t =
t
⎨ denotes the
⎩ 0, when low level is sent
waveform of an R-T command from the transceiver,
A denotes the amplitude of the RF signal transmitted
t
by the transceiver, A denotes the amplitude of the CW
emitted by the CWE, and φ () t is the phase different
between the two carriers. Let ∆ f = fc − ft
. After
some manipulations, Eq.(6) can be rewritten as Eq.(7).
s( t) = ⎡
⎣m( t) + Acos( 2π∆ ft+
φ( t)
) ⎤
⎦cos( 2π
ftt)
.(7)
+ Asin 2π∆ ft+
φ t sin 2π
f t
( ()) ( t )
The signal s()
t is firstly passed through the
envelope detection circuit inside the tag IC as shown in
Proceedings of the 2007 International Symposium
on Applications and the Internet Workshops (SAINTW'07)
0-7695-2757-4/07 $20.00 © 2007

Figure 3 [2]. Let s () t denote the output of the circuit
as presented in Eq.(8).
e
e
2 2
() = () + + 2 () cos( 2π∆ + φ()
)
s t m t A Am t ft t
.(8)
Utilizing the Taylor series, se
() t can be expanded as
Eq.(9).
2 2 1
⎡2Am () t cos( 2π∆ ft + φ()
t ) ⎤
Se
() t = A + m () t i{1 + ⎢
⎥
2 ⎢
2 2
⎣
A + m () t ⎥
⎦
2
1
⎡2Am() t cos ( 2π∆ ft + φ()
t ) ⎤
− ⎢
⎥ + remainders}
8
2 2
⎢
⎣
A + m () t ⎥
⎦
.(9)
se
() t is then passed through the LPF as depicted in
Figure 3. Assuming ∆f
is sufficiently large; the high
frequency terms are filtered by the LPF. The
output se
() t can be approximated as Eq.(10).
2 2 1
se
() t = A + m () t −
. (10)
2 2
4 A + m () t
As shown in experimental results, the R-T
command can be correctly recognized by Gen2 tags
under such a circumstance. Contradictorily, when ∆ f
is small, the R-T command is distorted by the CW
emitted by the CWE, and Gen2 tags do not respond the
command.
2.2. The modulation depth of R-T commands
In [1], the modulation depth (MD) of a valid R-T
command is specified to be greater than 80% and less
than 100% in terms of the incident electric field
strength at the tags. The MD is defined as Eq.(11).
EA
− EB
MD = , (11)
EA
where E
A
and E
B
denote the maximum and
minimum amplitudes of the RF envelope in an R-T
command respectively. Figure 4(a) and Figure 4(b)
illustrate an R-T command in an ordinary Gen2 system
and in the proposed system respectively. From Eq.(10)
and Eq.(11), the MD in our proposed system can be
2 2 2 2
A + A − A / A + A .For the
presented as ( t )
typical MD=90%, the power relation between the CWE
and the transceiver can be derived as Eq.(12).
t
P / d + P / d − P / d
' ' '
cw 1 tx 2 cw 1
'
'
Pcw
/ d1+
Ptx
/ d2
= 90% , (12)
'
where P
tx
denotes the maximum transmission power
of the transceiver while sending an R-T command.
2.3. The detection symbol in R-T Commands
In order to decode an R-T command, the tag IC must
have a PIE decoder, which is an edge detector [2]. A
regular Gen2 R-T command cannot be directly used in
the proposed system, because the first falling edge and
the last rising edge of the R-T command, as depicted in
Figure 4(a), cannot be detected by tags. In order to be
compliant with Gen2 tags, two short detection symbols
are added to the head and the tail of each R-T
command in the proposed system as illustrated in
Figure 4(b).
3. Experimental results
An Agilent 89601 vector signal analyzer
incorporating with an Agilent E4445A spectrum
analyzer is used to capture the communication signals
between the transceiver and a Gen2 tag. The
frequencies of the CWE and the transceiver are
915MHz and 917MHz respectively. The R-T Tari is
25us, and the duration of the both detection symbols
are 11us. The powers of the CWE and the transceiver
are adequately adjusted, so that the MD requirement is
satisfied. The snapshots are taken nearby the tags.
Figure 5 shows that the tag MBS is relatively strong.
Under such a circumstance, the accessible range of the
tag can be extended to 4 times longer than that of an
ordinary Gen2 system with the same emission power.
Figure 6 demonstrates an interesting multi-band
backscatter property of Gen2 tags. We let the
transceiver not only send the R-T commands but also
send a low power CW during the tag responses. We
found that the tag MBS presents not only in the carrier
frequencies of CWE but also in the carrier frequencies
of the transceiver (917MHz) as illustrated in Figure 6.
This phenomenon implies that we can apply the
frequency diversity technology (using a multi-carrier
CWE, for instance) to enhance the quality of the tag
MBS and to mitigate the multipath fading problem.
4. Conclusion
In this paper, we proposed a multi-carrier UHF
passive RFID System, which can effectively extend the
accessible range of Gen2 tags. Moreover, the multiband
backscatter property of Gen2 tags, which is
Proceedings of the 2007 International Symposium
on Applications and the Internet Workshops (SAINTW'07)
0-7695-2757-4/07 $20.00 © 2007

explored in this paper, can be utilized in future RFID
system design.
Acknoledgement
Amplitude
First falling
edge
Last rising
edge
E A
This study is partially supported by the National
Science Council of Taiwan under grant no.NSC94-
2218-E-011-011 and Yuen Foong Yu Paper MFG. CO.
References
[1] EPC TM Radio-Frequency Identity Protocols Class-1
Generation-2 UHF RFID Protocol for Communications at
860MHz-960MHz Version 1.09,
http://www.epcglobalinc.org, 2005.
[2] Zheng Zhu, “RFID Analog Front End Design Tutorial,”
http://autoidlabs.eleceng.adelaide.edu.au/Tutorial/RFIDa
nadesign.pdf, 2004.
[3] A Paulraj, R. Nabar, and D. Core, Introduction to Space-
Time Wireless Communications, Cambridge University
Press, 2003.
(a)
(b)
TAG TAG
P
tag
P
'
tag
d 1
d 0
d 2
'
P cw
CWE
P rx
P cw
Reader
'
P rx
Transceiver
'
P tx
Figure 1. (a) An ordinary UHF passive RFID
system. (b)The multi-carrier UHF passive RFID
system.
Amplitude
E A
(a)
Head Detection
Symbol
E B
E B
(b)
Figure 4. (a) The amplitudes of the RF envelope
in an R-T command in an ordinary Gen2 system.
(b) The amplitudes of the RF envelope in an R-T
command in the proposed system.
t
Tail Detection
Symbol
t
CWE
on
CW
off
Transceiver
Turn on
Query
…
Turn off
Gen2 tags
Figure 2. A simplified operation flow for the multicarrier
UHF passive RFID system.
t
Figure 5. A snapshot of the R-T and T-R
communications around 915 MHz, which is the
frequency of the CW emitted by the CWE.
Antenna
Rectifier
LPF
Hysteresis
Comapator
Envelope
Detector
Figure 3. The RF envelope detection and low pass
filter inside an UHF passive tag [2].
Figure 6. A snapshot of the R-T and T-R
communications around 917 MHz, which is the
carrier frequency of the transceiver.
Proceedings of the 2007 International Symposium
on Applications and the Internet Workshops (SAINTW'07)
0-7695-2757-4/07 $20.00 © 2007