Systematic design and prototyping of a low-cost passive UHF-RFID transponder

This paper presents a methodical design and prototyping of a passive European ultra-high frequency (UHF) band radio frequency identification (RFID) transponder. The transponder has a 70 × 17 × 0.3 mm 3 copper antenna whose design is based on the folded dipole architecture and utilizes techniques such as meandering and end loading to match a Texas Instruments (TI) UHF-RFID chip through a T-match feeding network. The tag’s simulated and measured performances indicate good coverage of the entire UHF band with a return loss better than 10 dB. The transponder was then fabricated using inexpensive off-the-shelf materials and its performance was tested. The proposed tag achieved good gain, read range, and cost efficiency when compared with current folded dipole antennas and can be easily adapted for various applications such as supply chain, access or security, and vehicle identification.

F I G U R E 1 RFID system operation UHF antenna. In passive RFID systems, reader-transponder communication is based on modulation and backscatter of EM waves, as depicted in Figure 1. In this operation, the reader generates and transmits a radio frequency (RF) signal through its antenna. A tag, located in the Fresnel or far-field zone of the reader, captures the impinging signal, and uses part of the harvested energy to power its electronic circuitry. The remaining portion of the signal is digitally encoded with the microchip's unique identifier code and the modulated signal is re-radiated. This modulated backscatter is then sensed by the reader's radio front-end, conditioned, and relayed to a control unit, where the transponder's identifier is determined. To efficiently extract power from the reader's RF signal, it is imperative that the tag's antenna impedance be a conjugate match of its chip's input impedance.
UHF-RFID transponder designs in literature vary largely in terms of antenna configurations, design approaches and complexities, as well as fabrication methods and materials. The design and manufacturing process of passive RFID transponders, which allows for effective operations within the UHF bandwidth as explained in the previous paragraphs can be costly and technical in terms of the materials used and the fabrication technique employed.
As most traditional RFID tag antennas are developed using etching technique, it is imperative to note that other design and fabrication methods like the printing technique exist. Chemical etching of thin copper or aluminium foils into the designed antenna pattern is expensive and hazardous 14 since the etchant (typically ferricloride) corrodes the metal foils with time. Thus, rendering the antenna less efficient and environmentally unfriendly when damaged. The printing technique which emerged because of the shortcomings in the etching method offers better advantages with varieties of technologies such as screen printing, 15,16 gravure printing, 17 inkjet, 18,19 flexography 20 and thermal transfer 21 process. This technique is currently employed in most electronic industries as it provides a fast and variable design options for the tag antenna to ensure efficient performance, and environmental friendliness. Hence, the transfer of electronic circuit designs onto substrate materials such as paper, plastic and fabric are made possible. RFID tag antennas developed using microstrip patch and inverted-F configurations for application on different surfaces (metal and curved) increases manufacturing complexity and cost. Thus, are limited to certain application and surfaces.
In most works, certain criteria are often traded-off and a few metrics are optimized to the detriment of these seemingly less important factors. It is, however, desired that all criteria be satisfied as possible. For example, in the work presented in Reference 22, the authors discussed a passive transponder whose antenna was modeled on the nested H-slot inverted-F microstrip configuration to improve read range. However, complex impedance matching, and tag miniaturization was quite challenging. This was partly because the tag relied on a non-resonant slot for impedance tuning. Moreover, the read range of the designed tag was quite poor, and the tag antenna required an external actor antenna to boost its read distance. This solution is not favorable in terms of cost, system simplicity, and ease of usability. In Reference 23, authors presented a design for a passive UHF-RFID tag with focus on flexibility and use for wearable applications. This work retrofitted a bow tie shaped resonating at 1.5 GHz for use in the UHF band. The down tuning was achieved by tracing the antenna arms with meander lines. The radiating body of the antenna was formed from aluminium tape whereas a mix of polycarbonate and polydimethylsiloxane (PDMS) materials were used as substrate. Although, the fabricated tag antenna performed satisfactorily in terms of read range and wearability, its dependence on special materials for fabrication does not make it an economical solution. The paper 24 proposes the design of a tag with a microstrip antenna etched on a FR4 substrate. It also employed a double T-match network coupled with folded dipoles. The double T-match configuration ensured effective matching of the transponder chip, enabled the reduction of differential mode noise, and allowed size reduction. However, the antenna gains and read range were limited due to the sizing of the antenna resonating structure. In Reference 25, the antenna was configured as a slotted dipole. The tag antenna was then realized through a special thermal ink transfer process. The design achieved good read range and robustness to EM interference of surrounding artifacts. However, the fabrication process is quite complex, expensive, and requires special materials and tools. In both References 26,27, UHF tags were developed by printing the antenna with silver (Ag)-based inks on polyethylene naphtalate and styrofoam respectively. The two works were aimed at producing flexible tags for various applications ranging from on-body sensing to conformal object installations. The tags achieved this goal, but the read ranges were poor, and the use of Ag ink is not economical. The novelty of this work lies in the design of a highly efficient yet economical passive tag antenna with 95% matching efficiency and UHF operational capability. The 70 × 17 × 0.3 mm 3 dimension of the tag antenna incorporated a folded dipole architecture and utilized meandering and end-loading techniques to match the UHF-RFID chip through a T-match feeding network. As opposed to similar folded dipole designs which achieved at most 80% matching efficiency as seen in References 27,38, it is imperative to note that this degree of efficiency was realized with large tag antenna dimensions of (81 × 20 × n/a mm 3 ) and (84 × 15 × 1 mm 3 ) respectively when compared to this work. Thus, approximately 16% × 15% × 70% miniaturization attained in this regard increases the scope of transponders application. Based on the −36.91 dB reflection coefficient recorded at 866 MHz frequency, a return loss better than the reference 10 dB was achieved in this work which implies a 23% improvement in the tag antenna's performance when compared to Reference 25. In contrast to its referenced counterparts, this work achieved about 50% savings in production cost per unit. This gain was due to certain key economic factors, which include material availability, fabrication time and cost per unit unlike in Reference 25,27 where expensive Ag-coated nylon yarn, polyimide and styrofoam substrates were used. These materials are not only expensive but market-wise inaccessible. The improvements discussed in this work delivered around 90% efficiency, which is beneficial when compared with similar works.
A summary of the vital attributes presented in Table 1 encapsulates the contributing areas and research gap achieved by this work.

Transponder design -target chip
The inlay of the passive UHF-RFID tag consists of an antenna designed for and attached to a Texas Instruments' RI-UHF-STRAP-08 chip. As shown in Table 2, the ISO/IEC 18000-6c compliant chip is a Gen2 IC (Integrated Circuit) mounted on two aluminium pads and operates in the 860 to 960 MHz band. The whole l st × w st = 8 mm × 3 mm strap assembly is affixed to a Polyethylene Terephthalate (PET) substrate. For a nominal read sensitivity, P th read,tag = −13 dBm, the die's input parallel resistive and capacitive values, obtained at 866.5 MHz using the load pull technique, are 440 Ω and 2.8 pF respectively. At mid-band, the strap poses an approximate 3.5 Ω resistance in series with a 0.095 pF capacitance. 28 However, for simplicity,

Design considerations
UHF tag antennas are often designed with a dipole-like structure due to the omnidirectional far-field radiation pattern of dipoles in the H-plane. Visually depicted in Figure 2, these antennas are generally long, thin, and possess two radiating arm that extend outward from a source (i.e., chip). In reader-to-tag communication, these dipole arms capture energy emitted from a reader's antenna. This energy builds up and causes current flow to the tag IC hence enabling far-field communication between reader and transponder. The geometrical dimensions of these arms determine the major properties of the antenna. The most used dipole structure is the half-wave dipole whose electrical length, at the desired frequency, approaches one-half of the wavelength, op , where the dipole exhibits maximum radiation. However, the physical length (∼17 cm at 866.6 MHz) is usually too long for most practical RFID applications. Additionally, neglecting losses and considering a very thin structure, the obtained 0.5 op dipole impedance at resonance is about ζ 0 /4π C in (2π) ≈ 73 Ω, where ζ 0 = 120π Ω is the free space wave impedance and C in is the alternate form of the cosine integral. 29 This impedance is farther from optimal for chip impedance matching. To achieve lengths for practical applications, the dipole must be made shorter. Shortening the dipole reduces the antenna resistance and makes the antenna capacitive. However, the antenna must have an inductive reactance to match Z chip for transfer of maximum power required for operation. The capacitive impedance of the shortened dipole may be tuned by introducing additional parallel 'counter-balancing' capacitance and/or adding series inductance. In most transponder antenna designs, both approaches are combined for a resultant complex matching. Matching can be achieved using techniques or structures such as end-loading, meandering, inductor-coupled loop, or the T-match feed. A combination of these methods is typically used as shown in Figure 3. End loading is accomplished by increasing the conductor area at certain parts of the dipole, most often, the outer ends. This conductor structure stores Tag antenna layout and dimensional parameters charges and creates a near-uniform current amplitude along the dipole. Thus, end loading is often employed in the design of tags for use on high dielectric surfaces. However, moderate end loading could be used to ensure that the tag is usable on a wide array of surfaces. Additionally, for a given dipole length and frequency, the increased conductor surface increases the resistance, R rad (equivalent to a cylindrical dipole with decreased length-diameter ratio). 29 Meandering is often used as a miniaturization technique where the antenna can be made to resonate at desired frequencies with reduced projected length. That is, the dipole length can be increased whiles the total antenna dimension is retained. This method involves the folding of the dipole arms. The folding introduces distributed inductances and capacitances essential for matching. Also, with sufficiently long meandered arms, the antenna can yield a gain like that of the straight half-wave dipole antenna. However, up to the first resonance of the antenna, currents on the adjacent meander line segments have opposite phases. These currents do not contribute significantly to the radiated power. Nonetheless, they produce ohmic losses. Therefore, R rad is mainly given by horizontal traces whiles adjacent vertical traces control electrical energy storage and ohmic losses. The loss in a thin, w d -wide, and h m -long trace made of material of conductivity σ and permeability μ, is attributed to the resistance R loss given in Equation (2).
The energy storage ability of a l sep -separated pair of such a trace is characterized by an inductance per unit length, l m and capacitance per unit length, R m given by Equations (3) and (4) respectively. 30 Thus, with meandering, the increase in average stored energy implies that resonance is achievable at much lower frequencies. The resonance frequency of a dipole with (periodic) meander lines could be estimated from the geometrical parameters of the meanders. As stated earlier, a meander could be represented as a pair of lines terminated with a short circuit. As derived in detail in Reference 31, the frequency at which a l ant -long dipole antenna with m meanders resonate, can be approximately deduced from Equation (5) (very accurate for m ≥ 14).

4
[ log where β is the propagation constant of the inlay medium (assumed vacuum in all cases hereafter). However, antennas with meander lines have low efficiencies and are often characterized by narrower bandwidths, BW, and particularly high sensitivity to frequency variations due to increased unloaded quality factor, Q. An alternative size-reduction method to F I G U R E 4 Coupled loop feed and equivalent circuit meandering that could improve antenna bandwidth and efficiency is the inverted-F antenna configuration or any of its variants. This configuration is often used in the designs where tags are intended to be attached to high-conductivity surfaces or to be hosted with other electronic structures. 32 Dipole antennas for RFID use are typically fed with an inductor loop. This structure, as illustrated in Figure 4, adds an equivalent inductance to the antenna. This introduced inductance is proportional to the shape or dimensions of the loop and the coupling between the radiating body and the loop. The strength of this inductive coupling, which can be modeled as a transformer, is dependent on the nature of the coupling section and can be tuned by varying the distance, d loop between the two structures. At the input of the loop, the resultant impedance seen is in Equation (6) is the quality factor of the radiating body near the resonant frequency, f o . The impedance of the dipole is Z d = R d (1 + j)V at the resonance. The total input resistance is related to the dipole resistance, R d = R rad + R loss and the transformer mutual inductance, M whiles the reactance, X in (f 0 ) depends solely on the self-inductance of the loop, L loop . The dependence of L loop and M on d loop and the loop shape or aspect ratios are discussed in analytical detail in Reference 34. Inductively coupled loops are very effective for use in designs for matching high Q chips or chips with high impedance phase angles. 32 The T-match is a matching technique, which can be used to transform the input impedance of a dipole type antenna. As a feed topology, it is often viewed as a fully-electrically coupled version of the inductive-coupled feed. The structure introduces short circuit stubs connecting the radiating arms to a shorter secondary dipole (with length l t ≪ length of main dipole l d ) placed in parallel at a distance, h t . The h t -separated traces with the short stubs, which serves as a step-up impedance transformer can be modeled as a folded cylindrical dipole. The two parallel conductors appear as a two-wire transmission line with characteristic impedance, Z 0 ≈ (ζ 0 /π) ln (h t / √ r d r t ). 29 The parameters r t = 0.25w t and r d = 0.25w d are the electrical equivalent radii of the stub and dipole respectively. 35 Together with the matching stubs, the two dipoles form a loop-like structure as shown in Figure 5. Since a folded dipole operates as a balanced system, by assuming that its current is decomposed into two distinct modes, this input network can be analyzed in the so-called antenna mode or the transmission-line mode. 29 The total impedance (fairly accurate for h t < < λ op ) seen at the T-match transformer input, Z in is the shunt combination of the transmission-line mode impedance and the transformed dipole antenna mode impedance. Z in is given by: where Z t ≈ jZ 0 tan(0.5 l t ) is the input (non-radiating) impedance of the T-match with current analyzed in transmission-line mode (characteristic impedance, Z 0 and with a short circuit load). 32  where γ ∈ (0,1]. In antenna mode, contrary to the transmission-line mode, charges are not reflected towards the input but rather "go around the corner" at the end. The antenna mode current in the arms of the folded dipole is split in a (1 + γ):1 ratio for resonant lengths. Therefore, Z d is transformed by a factor of (1 + γ) 2 to be in parallel with 2Z t . The geometrical parameters h t , l t ,w d and w t can be tuned in order to match Z in to Z chip . Matching charts could also be developed using Equation (7) to aid design. The antenna was designed as a single layer structure with total area, l ant × w ant = 70 × 17 mm 2 and to be fabricated from copper of uniform thickness t ant = 0.3 mm. The structure of the antenna was formed following the considerations discussed in Figure 3. Due to the low Q of the target chip (≈ − X chip R chip = 6.7), the tag antenna was fed using a T-match.
Additionally, this configuration guarantees near-maximum coupling and gives greater control of the impedance tuning process. The w d /w t ratio was also kept as low as possible to maintain matching agility. 32 Size reduction was achieved by the introduction of (periodic) meander lines. By varying degree of meandering (number, height, and spacing), the antenna resonance could be easily tuned to desired frequencies. However, the use of this technique was controlled to minimize ohmic losses. The antenna was also end-loaded to improve R rad . The dipole was loaded slightly such as to keep to a minimum, the capacitance variation that may arise due to the tag's eventual attached surface. Radiation resistance obtained from such end loading was often small. To boost R rad , the whole antenna dipole was designed as a folded dipole structure by introducing inwardly extending arms. As discussed before, at resonance, R d (mainly R rad ) can be improved as much as to a factor of 4, that is, for γ → 1 and l d → 0.5 op . Due to fabrication limitations, the width of all copper traces was constrained to values greater than 1.5 mm. Moreover, the antenna was designed as one connected structure. The tag was to be fabricated on a l sub × w sub = 135 × 48 mm 2 polyester transparency film substrate with dielectric constant r(sub) ≈ 4.1 − 5.2 and nominal thickness, t sub = 0.1 mm. However, since l sub , w sub < < λ op , far field losses due the substrate can be neglected.

Equivalent circuit model
The general objective of the design of UHF-RFID passive tags is to match the antenna impedance to Z chip to transfer maximum power to the IC. Unlike the HF RFID case, there are no definitive analytical formulas, which link the geometrical parameters of a passive UHF antennas to an electrical model. 36 However, some approximate models can be made to aid the design of the antenna. Referring to afore-mentioned considerations in Figures 4 and 5, the UHF antenna can be modeled as shown in Figure 6, at first approximation. The model could be further simplified as a series RLC lumped-element load with equivalent impedance Z in at frequencies near the antenna resonance. Additionally, if the antenna is symmetric across the feed midpoint, a balanced version of the resonant circuit can be split along the electric wall at the center and a half circuit, investigated.

Feed design
As an initial step in the design procedure, the feed structure of the tag antenna where the tag chip will be mounted was created. A primary function of the loop feed structure is to generate the bulk of an antenna input inductive reactance, X in that matches X chip at f op . This inductance, L feed is dictated by the shape and sizing of the feed structure. From the loop inductance model in Reference 34 and using a trace width, w d = 1.55 mm, an L feed ≈ − It can be observed from Figure 7 that the input impedance obtained from simulating the T-match bears fair agreement with the result, Z in ' obtained from the calculations, thus validating the T-match analysis. As foreseen, the feed structure produces the bulk of the conjugate of the IC's reactance, and this would be supplemented by the transformed impedance eventual dipole structure.

Dipole design
The dipole is designed to produce moderately high resistance required for matching since the T-match's resistance contribution is poor. One way to achieve this is to shift the antenna resonance point towards lower frequencies. However, Figure 8 indicates that the antenna resonates at a frequency (∼ 2 GHz) > f op . To successfully undertake this large shift, the dipole must produce a large, distributed capacitances and/or inductances but large enough in order not to offset the T-match reactance heavily after transformation. This was achieved by meandering the dipole arms. It should be noted that the desired reactance, X d was about a sixth of the test dipole X d '. Therefore, it was imperative that the distributed

F I G U R E 7
The T-match structure and input impedance plot

F I G U R E 8 Theoretical resonance frequency of l ant -long dipole antenna versus number of meanders
reactance of the meanders is effectively high inductive. Due to fabrication constraints, the minimum possible value for l sep was 3.10 mm and this was used as the initial value for meander spacing. This choice of l sep gave a good L m contribution, however C m was rather poor (see Equations 3 and 4). Thus, the 'bends' of the meanders are mostly inductive as desired. The inductive energy storage was reinforced by choosing h m as long as possible but enough to accommodate another antenna features such as folded dipole arm that were yet to be designed (i.e., h m ≤ w ant − 3w d ). The choice of w d and h m did not impact R loss heavily due to the very small skin depth of copper (δ ∼ 2 μm) in the frequency band. Using Equation (5), the resonance of the dipole as a function of the number of meanders was examined. For simplicity, periodic meanders were employed. Additionally, meanders were mirrored across the electric wall at the center of the dipole hence meanders were introduced in pairs that is, m = 2,4,6,8 … The theoretical relation between the resonance frequency and m is shown in Figure 8.
It can be seen that, the resonance frequency reduces as m increases. Due to the high Q of the T-matched test dipole and its possibility to grow further, a chain of moderate m meander lines was to be used for the design to keep final R in and X in in optimal value ranges. This assertion and the graph in Figure 8 were validated by simulations as seen in Figure 9.
From Figure 9, it can be observed that the introduction of the meanders shifts the resonance point of the antenna towards low frequencies along a near-straight line with gradient 2 L feed . This behavior can be inferred from Equation (6). Additionally, a growth of the antenna Q with increase in meander number can be seen in Figure 9. This is due to the rise in average stored energy while the power loss remains almost constant. With six meanders, the antenna resonates at around 1.23 GHz. m = 6 was chosen for the final design since reactance due to additional antenna structures and other parasitic could be exploited to move the resonance point closer to the f op . Moreover, the choice of m was constrained by the antenna dimensions. It should be noted that the loading positions of the meanders along the dipole length do not affect their contribution to the shift in the resonant frequency. 31 However, their interaction with other structures due to their positioning may be consequential. The end of the dipole was then loaded. Due to the antenna size and fabrication restrictions, the loading that could be introduced was not large hence it had very little impact on R d . However, the minor increase in R d causes X in to increase correspondingly. This is expected as an increase in the R d reduces the impact of the subtractive term of X in in Equation (6). Figure 10 shows the impedance values recorded for different end-loading lengths (l sb = 1.55, 3.05, and 4.55 mm). It is evident that the last bend into the end loading was inductive like the meander bends. This contribution lowers f 0 further to about 1.1 GHz. Moreover, the minimal variations in R in and X in even out and keeps the antenna Q fairly constant over the different l sb . Finally, folded arms were introduced to first boost R d (hence R in ) and increase the value of X in towards −X chip . The width of the end loading was increased to allow the attachment of the arms, creating more same phase currents which contribute to R d (see Figures 10 and 11).

F I G U R E 9
Resonance shift due to different meanders F I G U R E 10 Parametric study of end loadings effects on Z in and resonance

F I G U R E 11 Folded arm lengths Z in improvements
The impact of various folded arm lengths, l f (l f = 9, 18, and 27 mm) on Z in is studied in Figure 11. The impedance plot shows that for l f values less than the length of the meander chain, R in only increases slightly but experiences a large growth when l f exceeds the length of the chain. This is because the resistance increase arising from the folding is composed of an interspersion of the 'negligible' R loss (parallel conductor lengths) and R rad . The increase in X in is primarily because of inductance due to the fold bend.
These account for the larger resonance shift and increase in the antenna Q in the blue curve (l f = 27 mm) than the previous three plots. Like an x-oriented dipole, the antenna exhibits a toroid shaped radiation pattern with a directivity, D tag = 1.99 dBi in the E-plane as represented in Figure 13. In free space, the antenna radiates power accepted at its input port with an efficiency rd = −0.21 dB. Thus, the (broadside) gain of the antenna G tag is

Prototyping of transponder
The designed antenna was fabricated by cutting its outline from a t ant -thick flexible copper sheet using a Secabo® S60 vinyl cutter. The structure was then affixed to the cheap and transparent (polyester) substrate made from a projector F I G U R E 13 E and H principal plane cuts -3D radiation pattern F I G U R E 14 RFID transponder prototype transparency. The RFID die was mounted on the antenna by bonding its aluminium straps to the antenna's input port with lead-free solder flux. The prototype of the transponder is shown in Figure 14.

RESULTS AND DISCUSSION
The complex power reflection coefficient at the antenna inputs was measured and compared to the simulated values as represented in Tables 4 and 5. As shown in Figure 15, although the fabricated antenna exhibits a slight frequency shift, the measured and simulated S 11 bear fair agreement. Open space read range measurements were also performed to experimentally validate the performance of the tag prototype. The tag was positioned in the broadside of a 7 dBi (±1 dBi) left-hand circularly polarized (LHCP) reader antenna set on tripod stand. The reader antenna was excited/read by a CAEN A949EU RFID reader with maximum RF output power = 1.2 W through a 50 Ω coaxial line. 29 The reader hardware was also connected serially (via RS232 interface) to a measurement software on a computer. The reader's effective isotopically radiated power (EIRP), P eirp was set to 3.28 W. The measurement setup is shown in Figure 16. The variations obtained in the measured results of the prototyped transponder antenna as against the expected simulated results were due to the material properties (physical and chemical), and reader mismatch. Thus, these factors primarily accounted for the slight differences in the parameters considered. Assuming that both reader antenna and transmitter output port are perfectly matched to coax supply line and that the IC sojourns an equal fraction of time in each of its two load modulation states (with modulator Γ change, |ΔΓ| = 0.2 37 ), the maximum theoretical reading range at f op is where tag is the average of tag in the two modulation states and p m is the polarization mismatch factor. It should be noted that since the circular wave launched by the LHCP reader antenna is essentially two orthogonal linear waves, the LP tag antenna captures the in-phase component of the reader wave. Thus, for a given angular direction, p m = 0.5 irrespective of the transponder's orientation angle. The distance between the transponder and the reader antenna fixture was increased until the reader could no longer receive a modulated backscatter from the tag. The maximum-recorded read range was 1.72 m, which suffices as a low-cost solution for many RFID applications ranging from large item management in warehouses, access/security, and vehicular identification. The flexibility of the transponder makes it very adaptable for various platforms. The discrepancy in the measured read range of the prototyped transponder with respect F I G U R E 16 Read range measurement setup to the theoretical distance could be attributed to a couple of factors such as non-ideal properties of fabricated transponder antenna like G tag and Z in , modulation losses, chip damage (electrical or mechanical), reader mismatches or large path loss.
In a further study, the prototyping processes will be improved, and the transponder characterized in an anechoic chamber to accurately determine the limits of the transponder's properties. The features and performances of the transponder is compared with state-of-the-art UHF-RFID tag solutions in Table 6. The transponder presented in this paper performs fairly well against the state-of-the-art in terms of read range, second to Reference 25. The lower read range could be attributed to the slightly lower reader power used in our experiments as compared to that used in Reference 25. The transponder's performance although it has the smallest footprint, uses less reader power, and was realized using low-cost off the shelf materials.
Regarding the available references, the prototyped transponder's antenna possesses a good performance to cost ratio when compared with the state-of-the-art designs in Table 6. Based on the design materials, availability, cost, and fabrication processes, it can be deduced that the prototyped tag antenna developed in this work is more economical, and simpler to implement in achieving its operation in the European UHF RFID band.
In terms of its performance as against current designs, our prototype transponder achieved a similar read range to that recorded by Reference 25. Although 38 delivered a longer read range, it cannot operate effectively on metal surfaces due to significant interference. More so, the longest read range recorded by Reference 39 proved most effective when compared with this work nonetheless, it has no flexibility. Thus, the rigid nature of this fabrication limits its range of application unlike this work where a flexible substrate was employed. Therefore, it provides a wider area of applications.
Estimated cost efficiency is the ratio of estimated cost to estimated output expressed as a percentage. It is based on the efficiency formula and presented in Equation (14) as a manufacturing variant.

Estimated Cost Efficiency
× 100 (14) Overall Performance = ∑ (All Antenna Efficiencies ) From Table 7, 40 produced good results considering its cost efficiency and performance parameter. Nonetheless, results deduced from this work show that the transponder antenna can be easily prototyped, cost efficient and has better performance when compared with related works.

CONCLUSIONS
In this paper, the systematic design of a passive UHF-RFID tag has been outlined. Powered by the TI UHF Gen2 STRAP, the tag has a 70 × 17 × 0.3 mm 3 copper antenna affixed on a polyester substrate. The antenna covers the entire European UHF band with a return loss better than 10 dB in free space. The transponder was prototyped using inexpensive off-the shelf materials and its performance was tested. The designed tag had a small footprint and achieved good gain and read range. The measured performance fared well with a 12.5% increase in overall performance when compared to the state-of-the-art. With improved fabrication techniques, it can be mass-produced as a low-cost RFID transponder solution.