Microwave Journal
www.microwavejournal.com/articles/2784-circularly-polarized-aperture-coupled-patch-antennas-for-a-2-4-ghz-rfid-system

Circularly Polarized, Aperture-coupled Patch Antennas for a 2.4 GHz RFID System

The design of aperture-coupled patch antennas for an active read/write microwave tagging system using a circular polarization modulation

November 1, 1999

Circularly Polarized, Aperture-coupled Patch Antennas for a 2.4 GHz RFID System

Marcel Kossel, Hansruedi Benedickter and Werner Bächtold
Swiss Federal Institute of Technology (ETH), Zurich Laboratory for Electromagnetic Fields and Microwave Electronics
Zurich, Switzerland

Roland Küng
University of Applied Sciences Rapperswil (HSR)
Rapperswil, Switzerland

Jan Hansen
Swiss Federal Institute of Technology (ETH), Zurich Communication Technology Laboratory
Zurich, Switzerland

Antennas are key components of microwave RF identification (RFID) tagging systems because miniaturization of the transponder or tag is limited mainly by the antenna dimensions. Printed antennas are widespread in RFID systems due to their minimal processing costs. Compared to a classical patch antenna, aperture-coupled patch antennas offer an improvement in terms of the optimization of radiation characteristics and antenna size. The RFID system is operated in the 2.4 GHz industrial, scientific and medical (ISM) band. Due to the requirements of the modulation scheme, the aperture-coupled patch antennas are circularly polarized with switchable polarization sense.

This article presents the design of aperture-coupled patch antennas for an active read/write microwave tagging system using circular polarization modulation (CPM). First, a description of the tagging system setup is given that also shows the importance of antenna performance. Second, the design procedure for four different antennas is described with particular focus on a parameter study of antenna dimensions. Measurement results of the antennas' radiation characteristics that are relevant with respect to the modulation scheme are also presented. In addition, the performance of the RFID system is shown by means of an experimental setup using the presented antennas.

In recent years, there has been growing interest in the development of communication systems for the localization of objects.1,2 Examples of major applications for RFID or tagging systems are access control and security systems as well as identification systems for industrial automation. Most commercially available microwave tagging systems are operated in the 2.4 GHz ISM band.3 Only a few RFID systems use the 5.8 and 24 GHz ISM bands.4

In general, microwave tagging systems can be divided into three types. Remotely powered fully passive tags (type 1) must be operated in the near-field region of the interrogator antenna. Therefore, they are used only for short transmission distances.5 A second type of tagging system (type 2) uses battery-powered tags that show a lifetime of several years because the power supply is only used for the low frequency signal processing unit, consuming a few microwatts. Since there is no RF generator available on such a tag due to power economy, a passive reflex modulator is applied for the communication from the tag to the interrogator. Moreover, the loss of the passive reflex modulator converts directly into a reduction of transmission distance.6 A third type of tagging system (type 3) uses an active modulator on the tag to increase transmission distance. Since the active RF modulator consumes much more battery power than the baseband signal processing circuits, the decrease in battery lifetime must be weighed against an increase in transmission distance.

An RFID System Using CPM

The presented antennas are used in a type 3 RFID system that utilizes CPM. This modulation scheme is based on the principle that logical zeroes and ones are transmitted as left-hand circularly polarized (LHCP) or right-hand circularly polarized (RHCP) waves, as shown in Figure 1 . Basically, CPM is equal to on/off keying modulation of LHCP and RHCP waves.

Table I
Basic Interrogator and Tag Specifications

Interrogator

Effective isotropic radiated power (mW)

10

Frequency (GHz)

2.4 (ISM Band)

Modulation Scheme

CPM in combination wih FH technique (hopping rate 1250 hops/s

Tag

Power Consumption

standby < 100 µW
receiving < 500 µW
transmitting  < 3 mA

Transmission at interrogation

active backscatter 
(data rate 10 kbps)

Size (mm)

110/80/5

Block diagrams of the reader and tag are shown in Figures 2 and 3 , respectively, and a summary of system specifications is listed in Table 1 . CPM is used for the downlink communication for writing data to the tag. This modulation is performed by switching an RF carrier between two antenna ports for transmitting LHCP and RHCP waves. The tag antenna also shows two ports for receiving LHCP and RHCP waves. The received waves are separated and led to two identical RF detectors whose output voltages are sensed by a low power comparator. The comparator output signal corresponds to the demodulated data and is sampled by a microcontroller.

Backscatter modulation is applied to read data from the tag during the uplink communication. First, the reader transmits an unmodulated RF carrier as continuous LHCP waves. The received waves at the tag are frequency-shift keying (FSK) modulated and retransmitted as RHCP waves by an active modulator that consists of an RF amplifier and a mixer. The polarization sense is inverted for decoupling the input and output ports of the active device to assure stable operation of the active modulator. Since only one antenna is used at the tag, Rx/Tx switches select between the receive and transmit paths. At the interrogator, an in-phase/quadrature demodulator is used for the direct conversion during the uplink communication to prevent signal cancellation at critical distances (multiples of l/2) between the reader and the tag.

To generate the RF carrier at the reader, a frequency-hopping (FH) synthesizer is used. Frequency changes of the generated RF carriers are performed according to a pseudorandom pattern with a hopping rate of approximately 1.2 khop/s and a block size of eight bits per hop. The combination of CPM with FH techniques generates advantages in multiple-reader, multiple-tag environments where the main jamming threat stems from RF carriers of different readers at the same location. By randomizing the carrier frequency of each reader using FH methods, signal destructive fading and jamming can be reduced to short time intervals. Additionally, the use of spread spectrum techniques provides high immunity against multipath phenomena and frequency-selective slow fading channels.

CPM shows two advantages for microwave tagging systems. First, CPM allows the use of the total radiated power for the information-bearing signal. In comparison to other widespread modulation schemes in RFID systems (for example, FSK), no additional subcarrier is needed. Since the effective isotropic radiated power (EIRP) is limited in the ISM bands,7 an efficient use of the total radiated power with respect to the information-bearing signal is important to obtain long transmission distances. A second advantage of CPM is related to the low complexity of the demodulation circuitry on the tag. The demodulation is performed by comparing instantaneous power levels of incident LHCP and RHCP waves. Due to the comparison of relative power levels, a high dynamic range and improved sensitivity of the tag receiver are achievable. In addition, transmission errors due to linear-polarized jamming signals can be reduced because those signals are inherently suppressed by the demodulation circuitry.

These characteristics illustrate that CPM is well suited for low cost and power-saving active microwave tags. However, the combination of CPM and FH techniques for a microwave tagging system makes great demands on the radiation characteristics of the interrogator and tag antennas. First, the use of FH modulation requires broadband RF front ends with antennas that show large impedance and gain bandwidth. Because the transmitted data are circular polarization modulated, antennas with switchable polarization sense are demanded. In particular, the values for the axial ratio and cross-polarization isolation are of prime importance for the overall system performance. Cross-polarization isolation determines the separation of LHCP and RHCP, whereas axial ratio is a measure of circular polarization. The definitions of both antenna parameters as used in this article are shown in Figure 4 where the ratio of the magnitude of the orthogonal E-field components is the axial ratio and cross-polarization discrimination is defined as the ratio between received power levels at the LHCP and RHCP antenna ports received from an ideal circular reference antenna. A summary of the antenna specifications is listed in Table 2 .

Table I
Antenna Specifications

Center frequency fc (GHz)

2.44

Impedence bandwidth (%)

> 4

Axial ratio at fc (dB)

< 1

Polarization discrimination at fc (dB)

> 20

Antenna gain (dBi)

7

Circularly Polarized Antennas with Switchable Polarization Sense

Aperture-coupled patch antennas are especially suited to meet the antenna specifications because the multilayer structure of those antennas allows the radiation characteristics (patch layer substrate) and the size of the feed network (feed layer substrate) to be optimized independently. Figure 5 shows the structure of an aperture-coupled patch antenna with switchable polarization sense. The radiating patch is located on a substrate with relative permittivity er1 , and the feed network is etched onto the bottom side of a substrate with relative permittivity er2 . These substrates are separated by a common ground plane that features an electrically small slot aperture for efficient coupling of power to the patch.

The slot aperture shown comprises two orthogonal slots that are fed by two open-circuited stub lines of a 3 dB hybrid. This branch-line coupler produces fields of equal amplitude and 90° out of phase at its center frequency. Each input terminal of the hybrid provides an opposite circular polarization sense. The open-circuited stub length of the microstrip line extending beyond the slot aperture can be used for impedance matching and bandwidth enhancement. This stub, together with the aperture length, controls the input impedance over a wide range of values. Therefore, the lack of an additional matching network and the ability to choose a thin feed layer substrate help to minimize the size of the feed network. On the other hand, a thick patch layer substrate with low relative permittivity can be used to increase the impedance bandwidth.

Additionally, aperture-coupled patch antennas feature a high front-to-back ratio due to the intermediate ground plane. This characteristic produces good shielding of the RF front-end electronics (RF detectors, Rx/Tx switches and active modulator) and signal processing electronics at the rear side of the tag antenna. Although aperture-coupled patch antennas show increased manufacturing costs compared to classical patch antennas, their advantages in terms of radiation performance and antenna dimensions provide a viable alternative for microwave tag antennas.

This article presents the design of four different circularly polarized, aperture-coupled patch antennas with switchable polarization sense. The antennas' main differences are in the feed network, slot aperture shape and radiating patch. Three antennas are dual polarized and use a 3 dB hybrid as the polarizer in the feed network. Antenna type A is similar to the antenna shown previously. Its slot aperture consists of two orthogonal slots that are placed at some offset distance below the edges of the patch. Two pairs of orthogonal slots are used for antenna type B. A bent l/4 microstrip line is employed in the feed network of each pair to provide the required 90° phase difference between the orthogonal slots. Equal to a polarizer using a branch-line coupler, the polarization sense of antenna type B can be changed by switching between the two antenna ports. Antenna types C and D show annular aperture shapes. The annular square slot aperture of antenna type C is fed at two adjacent corners by the open-circuited lines of a branch-line coupler. Antenna type D is similar to antenna type C except that the radiating patch and slot aperture are circular. The stub lines feed the slot aperture as well in orthogonal directions. Figure 6 shows an overview of the four antennas. Figure 7 shows the top view on all layers of the antennas with switchable polarization sense.

Antenna Design

The four aperture-coupled patch antennas employ a 3.18-mm-thick substrate with er1 = 2.33 for the patch layer and a 0.508-mm-thick (20 mil) substrate with er2 = 2.22 relative permittivity for the feed layer. A planar electromagnetic (EM) solver8 using the method of moments was employed for simulations. In order to produce a good understanding of the mode of action of the relevant antenna design parameters, the focus is mainly on antenna type A. This antenna type can be considered a dual-polarized antenna that is fed by the decoupled output ports of a branch-line coupler. Since the design of a 3 dB hybrid is well known, only the design steps of the dual-polarized antenna are illustrated.

The dual-polarized antenna consists of two linear polarized elements, one of which is shown in Figure 8 along with its corresponding lumped-element equivalent circuit. Starting at the left end of the equivalent circuit, the slot aperture displays the characteristic impedance of the feed network Z0 . The slot aperture is modeled by a series inductance Lslot since the coupling between the slot and feed layer mainly occurs by the magnetic field component of the quasi-TEM mode propagating on the microstrip feed line. The radiating patch is modeled by a parallel resonant circuit where the resistance allows for radiation losses. The impedance of the radiating patch shows no reactive part at the resonance frequency. However, the series inductance of the slot contributes a reactive part to the input impedance that must be compensated by the capacitance of the open-circuited stub line in the feed network.

Based on the presented lumped-element equivalent circuit, three design parameters are used. First, the length of the open-circuited stub line can be varied to match the slot aperture to the impedance of the microstrip feed line. The size of the slot aperture and the offset distance from the slot aperture to the patch edge are two parameters that determine the amount of coupling.

Antenna Parameter Design Rules

Initial values for the dimensions of the radiating patch can be determined using equations of the modal expansion model or the open-circuited radial resonator model.9 Calculations for the patch layer substrate at the center frequency of the ISM band at 2.44 GHz produce a value of 38 mm and 18.5 mm for the antenna's patch length and radius, respectively. A feed line width of 1.56 mm is calculated to connect the dual-polarized antenna to the 50 W microstrip lines of the branch-line coupler. Figure 9 shows the initial antenna dimensions.

Figure 10 shows the behavior of the antenna's input impedance with changes in stub length of the feed network. As the stub length increases, the impedance locus rotates in a clockwise direction along a constant resistance circle at a fixed frequency. The input impedance curves on the Smith chart refer to different values of stub length, which differ up to ±15 percent from the nominal value of the equivalent circuit. The open-circuited stub length can be adjusted to obtain the desired reactance. For a 50 W impedance match,the stub length is adjusted until the input impedance at the design frequency is purely real.

While the stub length rotates the impedance locus, the aperture length controls the amount of coupling, as shown in Figure 11 . By increasing the aperture length, the input impedance curves move toward the right in the direction of resistance values higher than the characteristic impedance. The aperture length can be adjusted to obtain the desired resistive part of the impedance (for example, 50 W). Curves of different values of aperture length are depicted on the Smith chart.

A similar behavior of the input impedance occurs for slot width changes. For a narrow slot, the patch is undercoupled and the resonance resistance is less than the characteristic impedance of the feed line. An increase of the slot aperture also causes a reduction in the front-to-back ratio and thus reduces the patch efficiency. A small slot width of 0.6 mm shows the most effective coupling for the configuration. The slot aperture is operated below its resonance frequency.

The previous parameter study was focused on one of the linear-polarized elements. However, in the design of the dual-polarized antenna, it is important to obtain high isolation between the two feed lines. If the requirement of high isolation is not met, the coupler connection to the two feed lines results in a poor antenna axial ratio. The limited bandwidth of the coupler causes an additional deterioration of the isolation. Therefore, investigations of appropriate values for the offset distance of the slot aperture from the patch edge also must include the isolation between the orthogonal slots. As the isolation becomes higher with longer distance between the nearest corners of the slot apertures, a trade-off between aperture length and isolation must be made while increasing the offset distance of the slots. In the present case, these parameters are adjusted to obtain an isolation greater than 15 dB within the entire 2.4 GHz ISM band. For transmission of circular polarization modulated data, a high isolation is of prime importance as the ellipticity characteristic follows the isolation characteristic.

Although the presented design parameters are used mainly for the design of a type A antenna, some results for the design of the other antenna types also can be derived. The final dimensions of the presented antenna types are listed in Table 3 . Antenna types C and D with annular slot apertures show the same slot width and radial patch dimensions as antenna type A. Therefore, only the placement of the circular and square slot aperture relative to the patch edges as well as the length of the open-circuited stub of the microstrip feed line must be subsequently determined. For an optimal coupling of the configuration, it has been determined that the distance from the center of the patch to the annular slot is slightly less than half the radius of the patch for antenna type D or approximately a quarter of the length of the patch edge for antenna type C. The reactive part of the input impedance owing to the inductive slot aperture impedance can be tuned out as shown previously.

Table III
Antenn Dimensions

Antenna Type

A

B

C

D

Slot width (mm)

0.6

0.6

0.6

0.6

Slot length (mm)*

26

18

e=14

r=9

Offset of slot from patch midpoint (mm)

14.7

14.7

--

--

Stub length (mm)

14.0

22.3

11.0

11.2

Feed line width (mm)

1.56

1.56

1.56

1.56

Patch width (mm)

37

37

35

r=18.7

*e = slot length, r = radius

Antenna type B displays a different feed network. Two bent l/4 lines are used as polarizers instead of a branch-line coupler. The l/4 lines feed the orthogonal slot apertures 90° out of phase. Again, the open-circuited stub serves to match the first slot aperture. However, the reactance of the second slot must be compensated by a matching network that consists of a simple radial tuning stub.

Measured Radiation Characteristics

Small-signal scattering parameters have been measured within the frequency range of operation. A comparison of the measured S parameters with the EM simulations8 of the antennas (including the polarizers) showed deviations in the resonance frequency of less than 1.5 percent. This discrepancy may be caused by different substrate losses. For simulations, a loss tangent of 0.0002 at 2.45 GHz has been assumed for the RT/Duroid 5880 used as the feed layer substrate and a loss tangent of 0.0003 has been assumed for the Polyguide patch layer substrate. Figure 12 shows a comparison of simulated and measured SWR at the LHCP and RHCP ports of antenna type C within the 2.4 GHz ISM band. All antennas show an impedance bandwidth larger than four percent, which corresponds to the relative bandwidth of the ISM band. The specifications have been fulfilled because a thick substrate with low relative permittivity has been used for the patch layer.

Far-field measurements in an anechoic chamber have been performed to determine the antenna gain and radiation patterns. Figure 13 shows the measured radiation pattern of antenna type A in the horizontal plane. Helix antennas transmitting LHCP and RHCP waves were used as reference antennas to determine an antenna gain of 7 dBi. There are only small differences in gain among the four antenna types. The same measurement setup has been used to determine a front-to-back ratio of approximately 16 dB.

Measurements of the polarization ellipticity were performed in the near and far fields. In the far field, an automated electromechanical positioner was used to measure the axial ratio and polarization isolation. The antenna under test (AUT) is mounted on the positioner such that it rotates around its boresight axis. The received power levels at the left-hand and right-hand ports of the AUT are recorded while transmitting LHCP or RHCP waves from a purely circular reference source. If the reference polarization is specified by LHCP waves, power levels measured at the left-hand port are denoted as copolarized and power levels measured at the right-hand port are denoted as cross polarized.

Figure 14 shows the resulting plot of the measured power levels at the two antenna ports of antenna type A. The axial ratio is determined as the maximal aberration of the copolarized curve from a perfect circular curve. The polarization isolation is determined by the difference between the co- and cross-polarized curves. Although these measurements may differ from methods commonly used to determine the circular polarization (for example, the spinning dipole method10,11 ), they are well suited for measurements of the circular polarization relative to a well characterized reference antenna such as a helix antenna.

Near-field measurements have been performed by means of an automated E-field scanner12,13 for visualizing the circular polarization. The automated E-field scanner measures the magnitude of two orthogonal E-field components and displays them as the minor and major axes of an ellipse. Figure 15 shows the visualization of the circular polarization of antenna type A. The automated E-field scanner helps to visualize misalignments between patch, slot and feed layer, which produce an asymmetric distribution of the measured polarization ellipses across the radiating patch.

Table IV 
Measurement Results

Antenna Type

A

B

C

D

Impedance bandwidth (%) (SWR = 2)

8

7

6

6

Gain (dB)*

7.1

6.8

7.4

6.7

Axial ratio (%)

< 1.0

< 2.5

< 2.0

< 2.5

Cross-polarization isolation (dB)

> 1

> 8

> 12

> 10

*measured at 2.42 to 2.47 GHz

Based on the far-field measurement of the axial ratio and cross-polarization isolation, the values listed in Table 4 were determined. Within the measured bandwidth all antennas showed an axial ratio better than 3 dB. The best value of the axial ratio and cross polarization was obtained by antenna type A. As antenna types C and D use only one annular slot instead of two separated orthogonal slots, the isolation between their feed lines is reduced, which results also in a decrease in cross-polarization isolation. In addition, visualizations of the circular polarization with the automated E-field scanner showed that these antenna types are much more sensitive to misalignments between the different layers than in the case of antenna type A, possibly causing an additional deterioration of the axial ratio.

In order to minimize the size of the feed network, the l/4 lines of antenna type B are connected directly between two orthogonal slots instead of feeding the slots in an offset line configuration by using a power divider. Proper amplitude balance between the orthogonal slots is thus difficult to achieve, which leads to a degraded axial ratio compared to antenna type A. Due to the l/4 lines, the ellipticity bandwidth is slightly smaller compared to antenna types using a branch-line coupler as the polarizer.

The Experimental Setup of an RFID System Using the Presented Antennas

To demonstrate the superior performance of circularly polarized, aperture-coupled antennas for the presented read/write microwave tagging system, a fully operational prototype tag has been built using antenna type A. Figure 16 shows the front and rear sides of the prototype tag. At the rear side, the RF front end can be seen on the right part of the tag. It consists of a branch-line coupler for the antenna, two RF detector circuits and an active modulator as shown on the tag block diagram. The signal processing electronics on the left side also contain some test circuitry (data display and computer interface), which may be omitted for additional tag miniaturization. Due to the lack of an appropriate fixture, the substrates are tied together by plastic screws at a maximum possible distance from the patch and slot aperture. Off-the-shelf components have been used for the RF detector diodes, RF switches ( < 70 mW) and low power operational amplifier (PDC < 50 mW) of the comparator circuitry. Low barrier zero-bias Schottky detector diodes are employed for the RF detector circuits, which feature a voltage sensitivity greater than 10 mV/mW within the 2.4 GHz ISM band.

Since the ellipticity characteristic is important for the CPM, bit error rate (BER) measurements have been performed in the downlink from the interrogator to the tag. Starting with a straight alignment between the interrogator and tag, the tag was turned off in 10° steps during the measurement with a Wandel & Goltermann PF-4 BERT instrument. The resulting BER curve, normalized to the BER in the boresight direction, is shown in Figure 17 . There is only a slight decrease in BER within a rotation angle of ±30° from straight alignment. For rotation angles larger than ±70°, the received waves on the tag are much too depolarized, making demodulation by comparing LHCP and RHCP waves no longer possible.

Additional BER simulations were carried out to investigate the effects of the antenna's nonideal axial ratios. Starting with a perfect circular polarization of both transmitter and receiver, the axial ratio of the transmitter antenna was varied from 1 to 6 dB in 1 dB steps. The BER was simulated with respect to the signal-to-noise ratio (SNR) given as the quotient of the bit energy Eb at the receiver and noise level N0 due to the RF detector diodes and comparator. The curves  shown in Figure 18 were normalized to the BER simulation where an ideal axial ratio of 0 dB has been assumed for the transmitter antenna. For values of axial ratio up to 3 dB, no significant deviation in the BER of the perfect circular polarized antenna was observed. However, for higher axial ratio values, the performance of the system decreases sharply for high SNR.

Conclusion

Aperture-coupled antennas for a read/write microwave tagging system using CPM have been presented. The RFID system is operated in the 2.4 GHz ISM band. Due to the applied modulation scheme, circularly polarized antennas with switchable polarization sense are required. The interrogator also uses FH methods for randomizing the carrier frequency. Therefore, increased immunity against frequency-selective fading channels can be achieved. Additionally, different interrogators at the same location do not disturb each other when reading data from the tags because the shared use of a common carrier frequency is reduced to short time intervals. However, the combination of circular polarization with FH techniques is very demanding for a great number of antenna parameters, including the large impedance bandwidth, low axial ratio and high isolation between LHCP and RHCP waves.

The design of four different antenna types that differ mainly in the shape of the slot aperture and the polarizer in the feed network has been presented. A parameter study has been performed for a dual-polarized, aperture-coupled antenna type that uses a branch-line coupler as a polarizer. It has been shown that the most important design parameters are the open-circuited stub length in the feed network used as a tuning stub and the size of the slot aperture, which determines the amount of coupling.

In order to demonstrate the advantages of aperture-coupled antennas for the presented microwave tagging system, a fully operational prototype tag has been built using one of the investigated antennas. All RF front-end and signal processing electronics are placed on the rear side of the patch on the feed layer. It has been demonstrated that aperture-coupled antennas achieve a broadband radiation characteristic together with a small size of the feed network because the substrates of the multilayer structure can be chosen independently. In addition, aperture-coupled antennas demonstrate good shielding of the RF electronics due to the intermediate ground plane.

Acknowledgment

The work described in this article was supported by the Swiss Priority Program in Micro & Nano System Technology (MINAST). 

References

1.         U. Kaiser and W. Steinhagen, "A Low-power Transponder IC for High Performance Identification Systems," IEEE J. Solid-State Circuits, Vol. 30, March 1995.

2.         C.W. Pobanz and T. Itoh, "A Microwave Noncontact Identification Transponder Using Subharmonic Interrogation," IEEE Transactions on Microwave Theory and Techniques, Vol. 43, July 1995.

3.         http://rapidttp.com/transponder

4.         M. Kaleja, A. Herb, R. Rasshofer, G. Friedsam and E. Biebl, "Imaging RFID System at 24 GHz for Object Localization," 1999 IEEE MTT-S Digest, pp. 1497-1500.

5.         F. Carrez, R. Stolle and J. Vindevoghel, "A Low-cost Active Antenna for Short-range Communication Applications," IEEE Microwave and Guided Wave Letters, Vol. 8, No. 6, June 1998.

6.         T. Ohta, H. Nakano and M. Tokuda, "Compact Microwave Remote Recognition System with Newly Developed SSB Modulation," IEEE MTT-S Symposium Digest, 1990, pp. 957-960.

7.         IEEE Standard for Safety Levels with Respect to Human Exposure to Radio Frequency Electromagnetic Fields, 3 kHz-300 GHz (IEEE C95.1-1991).

8.         Momentum, HP Advanced Design System, 1998.

9.         I.J. Bahl and P. Bhartia, Microstrip Antennas, Artech House, Norwood, MA, 1980.

10.       J. Zürcher and F. Gardiol, Broadband Patch Antennas, Artech House, Norwood, MA, 1995.

11.       G. Evans, Antenna Measurement Techniques, Artech House, Norwood MA, 1990.

12.       T. Schmid, O. Egger and N. Kuster, "Automated E-field Scanning System for Dosimetric Assessments," IEEE Transactions on Microwave Theory and Techniques, Vol. 44, January 1996, pp. 105-113.

13.       Dosimetric Assessment System (DASY), Schmid & Partner Engineering AG, Zurich, Switzerland.

Marcel Kossel received his Dipl Ing degree in electrical engineering from the Swiss Federal Institute of Technology (ETH), Zurich, in 1997. In 1972, he joined the Laboratory for Electromagnetic Fields and Microwave Electronics of ETH, where he is currently working on his PhD degree. Kossel's current research is in the field of microwave tagging systems.

Hansruedi Benedickter received his diploma in electrical engineering from the Swiss Federal Institute of Technology in 1976. He has worked as a research assistant and senior research associate at the Microwave Laboratory of the Swiss Federal Institute of Technology and, since 1987, at the Laboratory for Electromagnetic Fields and Microwave Electronics. Benedickter's main research interests are microwave, millimeter-wave and on-wafer measurement techniques.

Werner Bächtold received his diploma and PhD in electrical engineering from the Swiss Federal Institute of Technology, Zurich, Switzerland, in 1964 and 1968, respectively. From 1969 to 1987, he was with IBM Zurich Research Laboratory. In 1987, he joined the Swiss Federal Institute of Technology as a professor of electrical engineering. Currently, Bächtold heads the Microwave Electronics Group at the Laboratory for Electromagnetic Fields and Microwave Electronics, which is involved in the design and characterization of GaAs MESFET and HEMT MMICs; InP-HEMT device and circuit technology; and modeling, characterization and application of semiconductor lasers.

Roland Küng (M '86) received his Dipl Ing degree from the Swiss Federal Institute of Technology (ETH), Zurich, in 1978. He was active in the design of HF communication systems from 1979 to 1983. In 1984, he founded a research group at Ascom Ltd. with interests in RF communications, digital signal processing and VLSI design. In 1992, Küng was named a professor at the University of Applied Sciences Rapperswil (HSR), Switzerland, where he teaches electronic communications systems and heads the wireless lab. He is also one of the owners of Elektrobit AG. Küng has special interests in the fields of RF architectures, spread spectrum techniques, algorithm engineering, software radios and RFID, and holds several patents on these topics.

Jan Hansen received his BSc in physics/mathematics from Trent University, Peterborough, Canada, and his Dipl Phys degree in physics from Freiburg University, Germany, in 1995 and 1998, respectively. He then joined the Communication Technology Laboratory at ETH, Zurich. Hansen is currently working on his PhD degree in the field of mobile communications.