See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/4047214

Behavioral Modeling and Simulation of Antennas: Radio-Frequency

Identification case study

Conference Paper · November 2003

DOI: 10.1109/BMAS.2003.1249866 · Source: IEEE Xplore

CITATIONS READS
10 1,286

4 authors, including:

Vincent Beroulle Rami Khouri

Grenoble Institute of Technology Sagemcom
114 PUBLICATIONS   583 CITATIONS    25 PUBLICATIONS   114 CITATIONS   

SEE PROFILE SEE PROFILE

Smail Tedjini
Grenoble Institute of Technology
459 PUBLICATIONS   3,490 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ANR LIESSE View project

RFID tag and systems View project

All content following this page was uploaded by Smail Tedjini on 27 May 2014.

The user has requested enhancement of the downloaded file.

 Behavioral Modeling and Simulation of Antennas :
Radio-Frequency Identification case study

V. Beroulle, R. Khouri, T. Vuong, S. Tedjini

LCIS - ESISAR
50 rue Barthélemy de Laffemas,
26902 cedex 9, Valence, France

Abstract using VHDL-AMS, the mixed-signal IEEE 1076.1 standard

Efficient simulation of wireless systems requires the modeling language [6].
development of antenna models compatible with
microelectronic tools. This article describes a first behavioral In the following part, a behavioral antenna model is described.
antenna model using the VHDL-AMS language. Moreover, we Then, in the third part, to validate this model and to demonstrate
present a complete behavioral RFID system model using this its interest, we propose a complete behavioral model of a Radio-
antenna model. Finally, several system simulation results Frequency Identification (RFID) system. A RFID system is
demonstrate the interest of this approach. described and major RFID design issues are discussed. In the
last part, several simulation results concerning these RFID
Introduction issues are briefly analyzed.
For many years, microelectronic market has been a constant
growing market. Predicted averaged growth rate keeps on high. Behavioral modeling of an antenna
This market is also a very competitive one. Thus, to conquer A. Transmitting mode
new application fields, designers must reduce the time-to- An antenna is “a means for radiating and receiving radio
market. This implies generating a first-time right design. waves”. According to [7], in the transmitting mode, an antenna
Moreover, to reduce costs and optimize performances (size, system can be represented by a Thevenin circuit equivalent as
power…), systems reach very high level integration. Success of shown in Fig. 1.
System-on-Chip (SoC) or System-in-Package (SiP), which are In this figure, the antenna is represented by an impedance ZA
embedding on the same chip or package several different digital given by :
and analog cores - microprocessors, memories, filters, Z A = RL + Rr + j X A (1)
amplifiers - confirms this major trend. Furthermore, progress of Radiating element is symbolized by a radiation resistance Rr
microelectronic processes allows the integration of new hybrid and an imaginary part XA. RL represents both the conduction and
components. Hybrid meaning that these components deal with the dielectric losses of the antenna. The source to which the
non-electrical signals : for example, Micro-Electro-Mechanical antenna is connected is represented by an ideal generator Vg
Systems (MEMS) or RF components. Obviously, the integration having its own internal complex impedance. The radiation
of these new hybrid components increases designs complexity. power delivered by the antenna is the power collected by
Designers need new models compatible with classical resistance Rr and it is given by :
microelectronic tools and adapted to microelectronic design
flow. 1 2
P r = I g Rr (2)
2
In this context of hybrid system integration, top level design Where Ig is the current through Rr.
verification of complete integrated Radio-Frequency (RF)
wireless communication systems, involving analog and digital Antenna
parts, is a major task. A hot topic concerns simulation tools for
these RF systems. In fact, the coexistence of a low baseband
RL
signal and a high carrier frequency involves high simulation
time. This problem is clearly addressed by RF simulation Vg
engines (Eldo-RF, Spectre-RF) specialized in non-linear system
simulations involving for example mixers or switches [1]. Due
to this new class of tools, efficient simulation of a complete Rg Ig Rr
integrated communication system is possible [2] [3] [4].
Nevertheless, there still is a lack of RF component models : few Xg
works concern RF signal propagation modeling or
electromagnetic effects modeling [5]. Moreover, this deficiency Xa
particularly concerns RF antennas modeling. Of course, without
antenna models, complete functional validation of wireless Source
communication systems is made impossible. Considering this Fig. 1 : Thevenin circuit equivalent of a transmitting antenna.
problem, this paper presents a first behavioral antenna model
Then, Ig magnitude is given by :
The equations extracted from both the receiving and
Vg transmitting antenna models allow us to describe the behavior
Ig = (3)
(R r + RL + R g ) + (X A + X g )
2 2 of a single bi-directional antenna.
This model permits to describe a transmission between two
Maximum power is delivered to the antenna under conjugate antennas using the VHDL-AMS language. This simplified
matching, which means when : model assumes no losses and no antenna mismatches.
R r + R L = R g
 (4) In order to show up the interest of this first behavioral model,
 Xa = − Xg
this antenna model has been included in a complete behavioral
If we consider a loss less antenna (RL=0), the ideal amount of model of a RFID system. The next part describes the major
power collected by Rr is given by: characteristics of this system.
2
Vg
Pr = (5) RFID system modeling
8 Rr
A. RFID overview
For a directive antenna, the radiated power per unit solid angle, Radio-Frequency IDentification (RFID) concerns technologies
also called radiation intensity, is given by: for tracking and access control applications [8]. RFID wireless
P r Dt systems allow non contact reading and are particularly effective
Ir = (6)
4π in environment where bar code labels could not survive. Market
Where Dt is the directivity of the transmitting antenna and 4π for these applications has been regularly growing since 1980.
describes a full surface solid angle. Theses systems generally consist of two parts :
- a base-station (BS) : a fixed transmitter and receiver.
Let us now consider a receiving antenna at a distance r from - a tag : a small mobile communication circuit stuck on
the transmitting antenna. The usually used parameter to tracked “products”.
describe the power capturing characteristics of this receiving The BS contains an antenna for RF signal transmission and
antenna is the effective area Ae given by : reception, and an electronic part consisting of an analog front-
λ Dr
2 end and a digital back-end.
Ae = (7) The tag, which is sometimes called “smart label”, consists of an
4π
antenna and an electronic part.
Where Dr is the directivity of the receiving antenna and λ its
operating wave length. RFID tags can be categorized as either active or passive. Active
tags are powered by an internal battery. Passive tags operate
The power collected by the receiving antenna (under far field without internal power source and get their operating power
condition) is given by : from the BS. These tags are consequently much lighter than
P r Dt Ae active tags. The trade-off is that they have shorter reading
P received = (8)
4π r 2 ranges than active tags. In our case study, we chose to model a
passive tag in order to deal with its design complexity.
B. Receiving mode Furthermore, RFID systems can be distinguished by their
Power collected induces a voltage Vt on the receiving antenna, operating frequency ranges. Low-frequency systems have
which is analogous to Vg of the transmitting antenna model. short reading ranges and low system costs. In these systems,
The Thevenin equivalent circuit of the receiving antenna and transmission only relies on the induction principle. High-
its load is shown in Fig. 2. The load to which the receiving frequency systems offer high reading speeds, long range
antenna is connected is represented by a complex impedance. transmission, and small antenna dimensions. For these
As we did for the transmitting model, we extracted the frequencies, transmission relies on RF signal propagation.
functional equations from this receiving antenna model. Compared to low-frequency systems, performances of high-
frequency RFID systems imply higher costs and more
complex design tasks. Thus, in the following case study
Antenna related to new tools for design flow optimization, we chose to
model a High-frequency RF system. More accurately we
chose a system operating within the 2.4GHz ISM (Industrial,
RL
scientific, and Medical) frequency band. The major advantage
of this band is to be available throughout most parts of the
Vt world.
It In the following, we first present the communication protocol
Rg
Rt
which has been described in our RFID system model. Then,
Rr we detail the front-end circuit for both BS and tag.
Xg
Xt
B. RFID simplified communication protocol
Xa Since our goal is to show the feasibility and the interest of our
antenna model for wireless system design, we have just
Load
implanted a simplified communication protocol. Nevertheless,
Fig. 2 : Thevenin circuit equivalent of a receiving antenna. this protocol is functional, even if it has been reduced
comparing to ISO/IEC 18000-4 RFID standard requirements.
This communication protocol, based on power activation and
two basic communication stages, is presented in Fig. 3.
 In the transmission chain, the baseband digital signal is firstly
Base-Station TAG
wait Manchester coded, then filtered with a low-pass filter. After
that, an ASK modulation is realized with a mixer working at
the local oscillator frequency fixed at 2.45GHz. Next, a RF
power amplifier adapted to the antenna is used to amplify the
request
P
T
R request
reception P
R
E
power of the signal before its transmission.
A

O N
S O C
E
In the reception chain, a low-noise amplifier is used before the
W M
I wait request
processing
W P
T
band-pass filtering for noise rejection. Then, a non coherent
E S
S E I
demodulation is realized to detect the signal. Finally, this signal
O
R I
O R N
is decoded and ready for use by BS back-end.
N data data
reception trans.

Tag front-end consists of a receiving chain equivalent to the

previous BS one and a “reflection chain” based on the
Fig.3 : Simplified BS - tag communication protocol. backscattering principle. In fact, backscattering is particularly
well adapted due to low-power requirements for tag
During the power up stage, the antenna of the BS (left part) operations. Moreover, this communication technique permits
transmits a RF signal to activate the tag (right part). Obviously, tag design simplification. Modulating the load impedance of
tag activation is possible only when tag passes through the BS the tag antenna in synchronization with a bit stream permits to
electromagnetic zone of influence. In this system, BS antenna / reflect or not the incident wave. Thus, this reflected signal is
medium / tag antenna are the conduits between the tag and the equivalent to a modulated signal coming from the tag.
BS.
Then, BS controls the system communication as illustrated in The front-end tag schematic is given in Fig. 5. All functional
Fig. 3. Firstly, tag detects the reader’s request signal : it blocks including antennas in both Fig. 4 and Fig. 5 have been
corresponds to the request and request reception stage. In the described using VHDL-AMS language. For the moment, only
tag and in the BS, this communication protocol is implemented high-level descriptions of these functions have been written.
in a finite state machine included in the digital electronic back- The following part presents several simulation results obtained
end. with these descriptions and discusses their interest.
Secondly, tag transmits a response, in our case its identification
number, to the BS : it is the data transmission / reception stage. TAG FRONT-END
Finally, the BS decodes these data previously encoded in the
c e f
tag’s integrated circuit and passes it to a host computer for Demodulation Manchester
A/D decoding
processing.
Power
d
C. Base-Station and tag font-end reception Energy
The BS radio front-end consists of two signal processing
chains : one for signal transmission and an other one for signal h g
Manchester
reception. A coupler, placed before the antenna, allows to drive Load
modulation encoding
the transmitted signal toward the antenna and the received
signal coming from tag toward the reception chain. Fig. 4 Fig.5 : Tag front-end.
presents a schematic view of these two chains. Components
involved in this BS radio front-end are filters (low-pass and Experimental results
band-pass), mixer, amplifiers (PA : Power Amplifier and LNA : The use of a behavioral system model allows designers to
Low Noise Amplifier)… In conformity to standard perform functional design validation earlier in the
requirements, these components realize an ASK modulation of a microelectronic design flow. This validation step permits to
Manchester coded digital stream. verify equivalence between original specifications and design.
The first simulation presented bellow checks the system’s
functionality, including several states of the communication
BASE-STATION FRONT-END protocol.
Furthermore, many evaluations of system performances using
a b
Manchester
encoding PA our behavioral system model are possible. For example, the
evaluation of the Signal to Noise Ratio (SNR) in different points
of the transmission chain can inform us about the
Local
oscillator communication quality. Thus, in the second simulation, we
propose to analyze the SNR of the signal received by the tag
k j i and the SNR of the signal received by the Base Station.
Manchester
decoding A/D demodulation LNA
A. Functional validation
Fig.4 : Base-Station front-end. We used Advance MS (ADMS) [9] from Mentor GraphiCs to
simulate our RFID system. This tool is a multilingual mixed-
signal mixed-mode simulator. The first work consists in
verifying our system architecture equivalence with the
previously described communication protocol. This simulation
considers a fixed distance between the BS and the tag and a
low-noise air medium.
Several signals obtained during the BS to tag transmission are
presented in Fig. 6. Then, signals for the tag to BS transmission, B. System performance evaluation
corresponding to backscattered signal, are given in Fig. 7. SNR is the usually used metric to evaluate a transmission
quality. In order to visualize the SNR, we chose to analyze the
eye diagram of the two received signals. ADMS provides a
special built-in function to automatically realize these diagrams.
Those eye diagrams are showed in Fig. 8.

The first eye diagram of Fig. 8 is related to signal c received

by the TAG. The second eye diagram of this same Fig. 8 is
related to signal i received by the BS after being
backscattered. The shape of the eye shows the effect of the
baseband filtering on the transmitted symbols. The eye
opening (aperture) shows how much the received signal is
perturbed by noise effects : the more the eye is open, better is
the transmission.
We can see on Fig. 8 that this aperture is much larger for the
Fig.6 : Base-Station to tag transmission. first transmission than it is for the second one. This means that
the SNR is much lower for the tag to BS transmission than for
In Fig. 6, the transmitted bit stream consists of the first two the BS to tag transmission. This is because of the double
stages previously presented in the communication protocol : propagation of the signal in the noisy medium.
- a power up stage at the beginning and the end of BS’s
communication.
In the case of a passive tag, the level of the received power
- a request transmission during 200µs.
becomes crucial. Thus, another interesting experiment would
In this figure, signal a contains the binary request (from 50µs
consist in evaluating the overall transmitted power from BS to
to 250µs). This request is then Manchester coded into signal b
tag depending on distance between them. This experiment
between two constant high level signals; constant signals
would permit to evaluate the maximum electromagnetic
corresponding to power up stages. Next, signal c is a
distance of influence of the BS.
continuous quantity representing the demodulated and filtered
signal received by the tag. A white Gaussian noise has been
added to this signal during its propagation in the medium ; d
signal is the power level of the tag ; e is the result of the
analog to digital conversion of c ; f is the decoded e signal, in
which we can recognize the original request.

Fig.8 : Eye diagrams of the signals received by the tag (c) and by the BS (i).

Conclusion
We presented a simplified behavioral model of an antenna.
Then, we included this model in a complete RFID 2.45 GHz
Fig.7 : Tag to Base-Station transmission. system behavioral model and we showed what this model
consists of. We also presented some of our simulation results
In Fig. 7, the transmitted bit stream consists of two stages in which the functional verification and the influence of noise
extracted from the communication protocol: on the quality of the communication. We showed how useful
- a power up stage. is the behavioral modeling of RF components, antennas in
- a data transmission / reception stage. particular, to mixed-signal integrated circuit designers, in
These stages are activated after the request reception stage. g order to validate their design’s functionality.
contains the binary identifier of the tag ; h is the Manchester We are currently working on a lower level of an antenna
coded signal. Then, i represents the received signal after model in order to provide designers with more accurate
demodulation and filtering. Finally, j and k are respectively the models. This antenna model will take into consideration
digitalized received and decoded signals. Obviously, k contains geometrical and technological parameters.
the binary identifier of the tag.

Signal analysis shows that our system architecture responds to

specification requirements. Despite of perturbations, request and
identifier are correctly transmitted through the medium. In the
following, we analyze the performance of our architecture.
 References
[1] K. Kundert, “Simulation Methods for RF Integrated Circuits”, San Jose,
California, Processdins of ICCAD’97 November 9-13, 1997.
[2] R. Ahola, D. Wallner, M. Sida, “Bluetooth Transceiver Design with
VHDL-AMS”, Deep Submicron Technical Publications, Mentor
Graphics, June 2003.
[3] P. Wambacq, G. Vandersteen, P. Dobrovolny, M. Goffioul, W.
Eberle, M. Badaroglu, and S. Donnay, "High-level simulation and
modeling tools for mixed-signal front-ends of wireless systems" book
chapter in Analog circuit design, structured mixed-mode design,
multi-bit sigma delta converters, short range RF circuits, edited by M.
Steyaert, A. van Roermund, J. Huijsing, Kluwer Academic
Publishers, pp. 77-94, 2002.
[4] R. Wittmann, J. Hartung, H. J. Wassener, G. Trankle, M. Schroter,
“Hot topic session : RF Design Technology for Highly integrated
Communication Systems“, DATE’03, Munich, March 2003.
[5] N. Milet-Lewis, G. Monnerie, A. Fakhfakh, D. Geoffroy, "A VHDL-
AMS library of RF blocks models", IEEE International Workshop on
Behavioral Modeling and Simulation (BMAS), 2001.
[6] IEEE 1076.1 workgroup, Language Reference Manual,
http://www.vhdl.vhdl.org/analog
[7] C. A. Balanis, "Antenna Theory, analysis and design", 2nd edition, John
Wiley & Sons, 1997.
[8] K. Finkenzeller, "Fundamentals and Applications in Contactless
Smart Cards and Identification" 2nd edition, Wiley & Sons LTD
April 2003.
[9] ADvance MSTM Reference Manual, Mentor Graphics,
http://www.mentor.com

View publication stats