RADIO LINK BUDGETS FOR 915 MHZ RFID ANTENNAS PLACED ON VARIOUS OBJECTS Joshua D. Griffin, Gregory D. Durgin, Andreas Haldi, Bernard Kippelen Georgia Institute of Technology School of Electrical and Computer Engineering Center for Organic Photonics and Electronics 777 Atlantic Dr. NW Atlanta, GA 30332-0250 ABSTRACT Passive radio frequency (RF) tags in the UHF and microwave bands have recently drawn considerable attention with their great potential for use in inventory management, parcel and postal tracking, as remote sensors, and in a host of other radio frequency identification (RFID) applications [1, 2, 3]. However, much more basic research is needed to increase the range and reliability of a passive RF tag s power and communication link, particularly when the RF tag is placed onto a lossy dielectric object or a metallic surface [4]. This paper presents the results of a radio assay, a series of tests for measuring the excess link loss of an RF tag antenna when placed on a variety of common materials. Using radio assay measurements for thin, flexible 915 MHz antennas, the results document the far-field pattern and gain change as the RF tag antenna is attached to cardboard, wood, acrylic, de-ionized water, ethylene glycol, ground beef, and an aluminum slab. It is shown that more than 18 db of excess loss on the backscatter link is added when RF tag antennas are placed onto lossy and metallic surfaces. 1. INTRODUCTION Passive modulated backscatter RF tags are transponders that communicate with an interrogator, or reader, using far-field electromagnetic waves. These tags operate in the UHF and microwave frequency bands and typically modulate the backscattered radiation reflected from the tag antenna using load modulation. In recent years, the use of passive RF tags for radio frequency identification (RFID) has generated increasing interest because of their potential applications to inventory management, parcel and postal tracking, access control, airport luggage, self check-out, and medical ID bracelets [1, 2, 3]. RFID applications of RF tags are expected to increase dramatically in coming years due to mandates issued by Wal-Mart and the Department of Defense for their suppliers to begin using RFID tags on their shipments [5, 6]. Although these entities desire to label and track individual items in their inventory, the relative high cost and poor performance of passive RF tags limit their use to tracking at the pallet or crate level [7, 8]. For item level tracking to become reality, much research is needed to lower the tag cost, decrease the tag footprint, and improve the tag performance and reliability. This material is based upon work supported in part by the STC Program of the National Science Foundation under Agreement Number DMR- 0120967, the Center for Organic Photonics and Electronics (COPE), the Georgia Research Alliance (GRA), and the Georgia Tech Foundation. 2. RF TAG SYSTEM PERFORMANCE RF tag performance is affected by many factors, including the electromagnetic properties of objects near or attached to the tag antenna. Although little literature has been published on this topic, Foster and Burberry [4] have recognized this issue and performed some basic measurements of an antenna near several objects. Raumonen et. al. [9] have studied a similar problem through simulation. Dobkin and Weigand [10] have also performed experiments that show a decrease in read range of several RF tags near metal and water along with the change in impedance of a tag antenna near metal. A convenient way to quantify the reduction in tag performance due to material attachment is as a decrease in the RF tag antenna gain, or gain penalty. A gain penalty measurement is a good figure of merit because it combines the effects of a decrease in the radiation efficiency, detuning, and antenna pattern distortion caused by material effects into a single number. The antenna gain penalty also provides a way to quantify the effects of poor antenna performance on the overall RF tag system performance through the use of two link budget template equations. The first equation (1) is a modified version of the Friis transmission equation that describes the amount of power received by the RF tag to operate the tag circuitry (assuming the antennas have an impedance and polarization match): P tag = P reader tx L sys + G reader tx (1) + G tag ideal Penalty 20 log 10 4π λ 20 log 10 (d) where P tag (db) is the power received by the RF tag, G reader tx (db) is the gain of the reader transmit antenna, L sys (db) represents the system losses in both the tag and reader, P reader tx (db) is the power input to the reader transmit antenna, G tag ideal (db) 4π is the RF tag antenna gain in free space, 20 log 10 λ is a frequency dependent loss, and 20 log 10 (d) is the free space path loss referenced to 1 m. If the gain penalty for a particular material is known, the free space RF tag gain can be adjusted to its actual value when attached to a material by substituting the gain penalty, in db, into Penalty in equation (1). The second link budget template equation describes the amount of modulated power that is backscattered from the RF tag to the tag reader. In the backscatter link, the RF tag receives a signal from
the reader transmit antenna, modulates it, and reflects a portion of it back to the reader receive antenna, shown in Figure 1. The twoway link budget is a modification of the monostatic radar equation defined by Balanis [11]. The key term in this equation is the radar cross section (RCS), σ, which determines the amount of power backscattered to the RF tag reader. The total RCS can be defined as follows [11]: σ = G2 tagλ 2 2 4π where G tag is the RF tag antenna gain assuming an impedance and polarization match, λ is the free space wavelength, and is a factor proportional to the reflection coefficient of the RF tag antenna load. By converting σ to the logarithmic scale and substituting it into the monostatic radar equation, the following two-way link budget template is obtained (all antennas are assumed to have an impedance and polarization match): P reader rx = P reader tx L sys + G reader tx (3) + G reader rx + 20 log 10 ( ) + 2 G tag ideal 2 Penalty 40 log 10 4π λ 40 log 10 (d) where P reader tx (db) and P reader rx (db) are the powers transmitted and received by the reader transmit and receive antennas, G reader tx (db) and G reader rx (db) are the gains of the reader transmit and receive antennas, 40log 10 (d) is the free space path loss for the two-way path referenced to 1 m, and all other terms are as defined in (1). As before, if the appropriate RF tag antenna gain penalty is known, it may be substituted into Penalty in (3) to account for the decrease the RF tag antenna gain due to material loss. Note that in (3), G tag ideal and Penalty are larger by a factor of two when compared to (1). Fig. 1. An RF tag system far-field backscatter communication link between a reader s transmit and receive antennas and an RF tag. 3.1. The Radio Assay 3. METHODS To measure the RF tag antenna gain penalty incurred when an RF tag antenna is placed on an object, three RF tag antennas were put through a series of tests, or radio assay, on various materials with differing permittivities and loss tangents. The term assay is taken from chemistry and refers to a series of tests applied to an unknown material in order to determine its properties. Hence, the radio assay (2) is a series of tests to determine the unknown gain penalties of the seven materials. The equipment used for the radio assay includes the following: a transmission line topology transformer to feed a planar RF tag antenna using a coaxial transmission line, a balun to eliminate spurious radiation at the antenna feed, a tunable impedance transformation network to achieve a conjugate impedance match between the RF tag antenna and its feed, directional antennas, an antenna range, RF tag antennas, a spectrum analyzer, and an RF signal generator. 3.2. RF Tag Antennas The RF tag antennas tested in the radio assay, shown in Figure 2, were planar folded dipoles designed for 915 MHz, a common frequency used for RFID applications. Each RF tag antenna was designed for an estimated input impedance of 233Ω. Two of the RF tag antennas were made from electroless silver and electroless copper deposited on flexible, 25.4 µm thick polyethylene terephthalate (PET) substrates. The conductor thicknesses of these antennas were 202±19nm and 195±17nm while their ohmic losses (measured at DC) were 36Ω and 49Ω, respectively. Since these prototype antennas were very thin and lossy, a baseline antenna was also tested as a point of comparison. The baseline antenna was milled on 1 oz. copper-clad FR4 and had negligible ohmic losses. Fig. 2. A planar folded dipole RF tag antenna designed for 915 MHz with an estimated input impedance of 233Ω. 3.3. Radio Assay Test Setup The gain penalty for an RF tag antenna attached to a material is the difference between the linearly averaged free space region antenna gain, measured in free space, and the same gain measured when the antenna is attached to a material. The linearly averaged free space region gain was calculated by averaging, in the linear scale, the power received in the free space region (i.e. the region spanning 270 to 90 shown as the shaded regions in Figures 7-11) from a horn antenna and solving the Friis transmission equation for the RF tag gain. To measure the received power, the RF tag antenna was used to transmit a +16 dbm, continuous wave (CW) signal at 915 MHz which was received by a horn antenna, as shown in Figure 3. A spectrum analyzer was used to measure the received power and was set to a center frequency of 915 MHz with a resolution bandwidth of 10kHz. Since two of the RF tag antennas were made from thin conductors with high ohmic losses and poor manufacturing tolerances, it was expected that their input impedance would differ from 233Ω. Hence, a tunable impedance matching network was used to achieve a conjugate impedance match between the RF tag antenna and the 50Ω coaxial feed line. In each test, the measured return loss was in excess of 46 db. However, since the impedance of the RF tag antenna is very low when attached to metal, it is likely that the tunable impedance matching network may have reached its tuning limits. All other devices were matched to 50Ω and the RF tag and horn antennas were oriented for a polarization match. Before each set of measurements was taken, the losses in the radio assay test setup were measured by
removing the horn and RF tag antennas and measuring the losses in the TX and RX portions of the system. The measurements were conducted in an outdoor antenna range that was free from scatterers and large enough for far-field measurements at 915 MHz, shown in Figure 5. The RF tag antennas were fed using a quarter wave coaxial balun connected to a coplanar strip (CPS) transmission line, as shown in Figure 4(a). Each RF tag antenna was connected to the CPS line using a clamp to press the antenna against the two conducting strips of the CPS line. Every effort was made to reduce coupling of the antenna to surrounding objects: the clamp fixture used to hold the antenna was made from unplated FR4 and nylon, the quarter wave coaxial balun connected to the CPS line was only 0.03λ o in diameter and was oriented orthogonally to the plane of the pattern measurements, the metal box housing the tunable impedance transformation network was separated from the antenna by 1.25λ o, and the antennas and materials were suspended on a PVC mount for tests as shown in Figure 4(b). (a) (b) Fig. 3. The radio assay experimental setup for measurement of the RF tag antenna gain penalty when attached to various materials. Fig. 4. (a) The feed of the silver RF tag antenna composed of a quarter wave balun, a CPS transmission line, and a clamp fixture. (b) The PVC antenna mount showing an RF tag antenna attached to wood, the quarter wave balun and the tunable impedance matching network. 3.4. Radio Assay Materials Seven materials, excluding free space, were tested in the radio assay. These were cardboard (several pieces layered to form an approx. 31 x 65 x 4 cm stack), wood (three pieces of pine plywood layered to form an approx. 30 x 57 x 3 cm stack), acrylic (six sheets of plexiglass layered to form an approx. 15 x 61 x 3 cm stack), de-ionized water (held in an approx. 22 x 10 x 10 cm waxed cardboard carton with a 0.6 cm cardboard backing), ethylene glycol (undiluted anti-freeze held in the same container as the water), ground beef (the meat was formed into an approx. 11 x 22 x 1.0 cm block in a thin plastic bag and taped to a thin piece of cardboard), and an aluminum slab (two pieces 22 gauge aluminum sheet metal were bolted together to form an approx. 46 x 58 cm sheet). These materials were chosen because they are in common use, are likely candidates for RF tag attachment, and provide a variety of complex permittivities and conductivities. The material properties of these materials were estimated from the permittivity and loss tangent values of similar materials given by Von Hippel [12]. The values at 915 MHz were interpolated from Von Hippel s data at 300 MHz and 3 GHz and are given in Table 1. The RF tag antennas were attached to each of these materials using thin adhesive tape. In the case of the liquid materials, the antenna was taped to the carton holding the liquid. Due to the CPS transmission line feed and the clamp fixture, the antennas were not able to lie perfectly flat against each material. However, since the distance that they were separated from the material or bent to touch the material was electromagnetically small (1/50λ o), the antennas were considered to lie flat against the surface of each material. 4. RADIO ASSAY RESULTS Figure 6 shows the measured free space patterns of the three RF tag antennas. From these patterns, the free space gains of the silver and copper antennas, relative to the baseline antenna, were calculated to be -1.9 db and -2.3 db, respectively. Note that these pattern measurements closely resemble those of an ideal dipole: the patterns contain no spurious lobes, are free from the effects of multi-path fading, and show no pattern distortion from cables or surrounding metallic objects. Figures 7-11 show the measured patterns of the antennas attached to the radio assay materials. Clearly, these materials induced pattern distortion which varied significantly with the material. The antenna patterns measured on cardboard (not shown) were almost identical to the free space patterns shown in Figure 6. The acrylic and wood patterns both displayed a pattern that was slightly larger in the material region than in the free space region. This increase is attributed to an increase in the near field power coupled into the material due to its high permittivity relative to air. The patterns measured on metal, ethylene glycol, and ground beef all showed a decrease in
Table 1. The average gain penalty on each material with the corresponding interpolated permittivities and loss tangents. CB Acrylic Wood Water EG Beef Metal Average Gain Penalty (db) 0.9 1.1 4.7 5.7 7.4 7.4 9.4 ǫ r - 2.6 1.7 77.3 33 50 - tan δ - 0.0061 0.036 0.048 0.4 0.7 - CB = Cardboard, the permittivity and loss tangent was assumed to be that of free space EG = Ethylene Glycol This paper presented a radio assay to measure the gain penalty of three folded dipole RF tag antennas attached to seven different materials. The measured gain penalties increased with the increasing loss tangent of the material from 1.1 db on cardboard to 9.4 db on metal. According to the two-way link budget template equation, the gain penalty due to metal would cause more than 18 db of excess loss in the backscatter link. Antenna pattern measurements were also presented that showed significant distortion. For the high loss materials (i.e. water, ethylene glycol, ground beef, and metal), the patterns became more omnidirectional while for some of the dielectric materials (i.e. wood and acrylic), the patterns showed an enlargement in the material region. Fig. 5. The antenna range used for the radio assay test. the nulls and a generally more omnidirectional pattern. The patterns on water were unique, with three distinct lobes centered at 0 and a pattern enlargement in the material region. The changes in the patterns on each material cannot be attributed to a single phenomenon. Surface waves, diffraction, the permittivity of the material, and the loss tangent of the material all have an effect. Note in Figure 8 and Table 1 that the copper RF tag antenna was not tested on metal. Table 1 shows the average RF tag antenna gain penalties. These values were calculated by linearly averaging gain penalty measurements for material. As Table 1 shows, the average gain penalty increases with the increasing loss tangent of each material. The individual measurements showed high precision for the dielectric materials with a small loss tangent (i.e. cardboard, acrylic, wood, and water), varying at most by 1.3 db. However, for the highly conductive materials, the measured gain penalty for each antenna varied by 1.8 db, 8.3 db, and 4 db on ethylene glycol, beef, and metal respectively. The values from Table 1 in conjunction with equations (1) and (3) show that the decrease in antenna performance due to material loss can be significant. For the backscatter link, (3), the measurements show that an antenna attached to metal will decrease the tag antenna gain by 9.4 db and produce an overall 18.8 db loss in the link. Fig. 6. The measured E-plane antenna patterns of each 915 MHz half-wave folded dipole in free space. 5. CONCLUSIONS Fig. 7. The measured E-plane antenna patterns of each 915 MHz half-wave folded dipole attached to wood.
Fig. 8. The measured E-plane antenna patterns of the baseline and copper 915 MHz half-wave folded dipole attached to metal. Fig. 10. The measured E-plane antenna patterns of each 915 MHz half-wave folded dipole attached to ground beef. Fig. 9. The measured E-plane antenna patterns of each 915 MHz half-wave folded dipole attached to ethylene glycol. Fig. 11. The measured E-plane antenna patterns of each 915 MHz half-wave folded dipole attached to water. 6. ACKNOWLEDGMENTS The authors would like to thank Yenpao Lu and Joel Prothro for their assistance in these measurements and Chris Durkin for his review of this paper. 7. REFERENCES [1] K. V. S. Rao, An Overview of Backscattered Radio Frequency Identification System (RFID), in 1999 Asia Pacific Microwave Conference, vol. 3, pp. 746 749. [2] I. D. Robertson and I. Jalaly, RF ID Tagging Explained, Communications Engineer, vol. 1, no. 1, pp. 20 23, 2003. [3] K. Finkenzeller, RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification, John Wiley and Son LTD, New York, 2nd edition, 2003. [4] P. R. Foster and R. A. Burberry, Antenna Problems in RFID Systems, IEE Colloquium on RFID Technology, pp. 31 35, 1999. [5] T. Purdum, Factory to Foxhole: RFID Deadline Looms, Industry Week, vol. 253, no. 11, pp. 12, 2004. [6] S. Ricca, Coming year critical for RFID, Official Board Markets, vol. 80, no. 47, pp. 1 4, 2004. [7] Anonymous, Wal-Mart RFID Deadline Won t Be Met, Official Board Markets, vol. 81, no. 1, pp. 4, 2005. [8] S. Ashley, Penny-wise Smart Labels, Scientific American, vol. 291, no. 2, pp. 30 31, 2004. [9] P. Raumonen et. al., Folded Dipole Antenna Near Metal Plate, in IEEE Antennas and Propagation Society International Symposium 2003, vol. 1, pp. 848 851. [10] D. M. Dobkin and S. M. Weigand, Environmental Effects on RFID Tag Antennas, in IEEE International Microwave Symposium, June 2005. [11] C. A. Balanis, Antenna Theory: Analysis and Design, John Wiley and Sons Inc., New York, 2nd edition, 1997. [12] A. R. V. Hippel, Dielectric Materials and Applications, The Technology Press of M.I.T. and John Wiley and Sons, Inc., New York, 1954.