A New Chipless RFID Permittivity Sensor System

Permittivity sensing is commonly utilized in multiple fields for various applications. In this study, an enhanced RFID (Radio Frequency Identification) permittivity sensor system is designed. Additionally, we investigated the signal processing procedures in a stepwise manner to obtain the resultant graphs. For validation, a simplified system including a reader and a tag is tested. The reader in the system comprises a pulse generator and a pair of antennas connected to an oscilloscope, whereas the tag comprises an antenna and a delay line sensor. The pulse generator produces short pulses of less than 1 ns periodically, whereas the short-ended delay line on the tag acts as a sensor, wherein the delay time of the traversing wave is primarily associated with the relative permittivity of the material under test (MUT). We tested the system to derive an equation for the value of relative permittivity of the MUT using the measured signal. The system was calibrated by performing both unloaded and air tests to obtain measurement equations. Moreover, an additional test validated the water permittivity sensing, and the relative permittivity measured was 78.1. Considering the errors affecting the measurements, the measured relative permittivity of water concurs with the actual value.


I. INTRODUCTION
Sensors are commonly used to measure different properties of a material. Among the multiple types of sensing techniques, permittivity sensing is applied in various fields, including biomedicine, agriculture, architecture, and automobile manufacturing [1]- [4].
Several methods, such as the transmission line method, resonant technique, open-ended coaxial probe method, timedomain reflectometry (TDR), and free-space method, are used to test the permittivity [5]- [8]. In applications such as safety and engine monitoring system, portable measuring devices that can collect information in real-time are highly preferred. Generally, the laboratory equipment (e.g., testing probes) required to monitor the materials are extremely bulky and warrant high implementation cost. Therefore, several studies suggested wireless testing and reported structural health monitoring applications using radio frequency identification (RFID) sensors [9]- [12]. In the case of RFID sensors, the reflection spectrum can carry substantial information The associate editor coordinating the review of this manuscript and approving it for publication was Weiren Zhu . including permittivity. However, the limited detecting range must be designed specifically based on the requirements. For instance, a remote testing technique was designed for corrosion characterization based on frequency selective surface design and feature fusion to generate resonances and enhance the sensitivity and reliability of the system [13].
This study investigated the feasibility of a chipless RFID permittivity sensor. The operating frequency of the sensor was set to 2.45 GHz, which lies within the industrial, scientific, and medical (ISM) frequency band. We applied and tested the remote monitoring technique for ease of control in water, verifying the ability of the proposed system to be applied in various environments.

II. THEORY
The proposed system is implemented based on RFID technology using a transmitter, tag, and receiver/monitor. The tag is attached to or dipped in the material under test (MUT) depending on the type of material. Additionally, the tag includes a sensor that translates the permittivity of the MUT into an electrical signal. The receiver recognizes this signal and displays the result on a monitor. Fig. 1   block of the RFID permittivity sensor. For simplification, no amplifiers are included in the tested system.

A. SYSTEM OVERVIEW
As illustrated in Fig. 1, the pulse generated from the source passes through the antenna before being transmitted through the air. The transmitting antenna is a microstrip antenna with a filtering characteristic at 2.45 GHz. The receiving antenna awaits the response signal that consists of two parts. The first part of the signal is directly coupled from the transmitter (V 0 ), whereas the second part is reflected from the tag (V r ).
As the receiving and transmitting antennas are adjacent to each other, the transmission time t 0 between the two antennas is considered to be 0. A 6-GHz oscilloscope with a sampling rate of 20 GSa/s acts as a reader to display the waveform of V 0 + V r . The delay time between V 0 and V r measured by the oscilloscope in the reading range indicates the total travel time of V r , calculated using (1).
where t 1 and t d denote the time spent by the signal traveling in air and the time delay generated in the tag, respectively. The sensor implemented in the tag generates t d , and the permittivity of the MUT affects the group delay of the wave in the sensor. Therefore, the measured t d is directly associated with the MUT permittivity.

B. DESIGN OF THE TAG
The tag constitutes the key component in the system, comprising a permittivity sensor and a broadband monopole patch antenna to maintain a stable received signal. The permittivity sensor is the core of the proposed RFID system, as the reading cycle of the sensor affected by the MUT permittivity is converted into an electrical characteristic, which in this case is the time delay. Fig. 2 illustrates the schematic of the delay line sensor based on the type of delay line selected for the experiment. Typically, the delay line must be long enough.
The sensor uses a pulse wave rather than a continuous wave. The pulse wave generated at the input port of the sensor in the tag antenna travels through the delay line within a certain time. Subsequently, the reflector directs this wave to the input of the delay line. The delay line can be open-ended, short-ended, or terminated by a capacitive load, wherein each  of these types differs in terms of phase. We used the shortended reflector in our experiment owing to its optimal performance with minimal noise. The pulse reflected from the end of the delay line travels into the antenna as a reflected signal. Further investigations are needed to calculate the total time spent by the pulse in the delay line.
The structure of the delay line is selected considering that the field generated on the transmission line can penetrate the MUT, which affects the transmission line characteristics [14]. Moreover, the loss should be sufficiently small for a long transmission line. Thus, the grounded coplanar waveguide (GCPW) is suitable for the delay line application [15].
where ε 0 denotes the permittivity of air and ε 1 and ε 2 represent the values of relative permittivity of the substrate and metal conductor, respectively. (2) suggests that ε eff is proportional to the ratio of h to b irrespective of the ratio of S and W [15]. This implies that the widths of S and W must be narrow to achieve maximum effective permittivity.
The total time (t x ) required for a pulse to traverse through GCPW in a single direction can be calculated using (3).
where l denotes the physical length of the GCPW and c is the speed of light (3 × 10 8 m/s). As the pulse travels a round-trip  in our design, the time (t d ) required for the signal to enter the delay line and return is determined using (4) The delay time is directly associated with the delay line structure and MUT permittivity. The transmission line-type delay line must be sufficiently long in a certain area. Furthermore, we did not consider the characteristic impedance as the designed antenna matches the impedance of the feed line. However, in the case of self-designed printed circuit boards, the transmission line width (W ) and gap (G) must be practically feasible. Figs. 4 and 5 depict the design layout and details of the delay line used in this study.
We fabricated the sensor on a Teflon substrate using the etching process. The total area of the sensor is 50 mm×50 mm with the line width and gap as 0.2 mm for ease of fabrication; a gap (G a ) of 0.3 mm is maintained between two lines. We performed a transient simulation using Ansys HFSS 19.2 to simulate a pulse traveling through the delay line [16]. The ADS  2016.01 software determined the characteristic impedance as 105 , which concurred with the simulated results. Additionally, the one-way delay time t x1 was calculated as 1.41 ns. The light blue region in Fig. 6 indicates the water layer that covered the sensor, in which the permittivity of water is set to 78.9 [17]. The simulated one-way delay t x2 was obtained as 12.61 ns. Fig. 7 depicts the avalanche pulse generator circuit adopted from the application note on 2N2369 transistor by Jim Williams [18]. It generates narrow-width pulses of 750 ps from the reader transmitter (Fig. 8). The transmitter antenna is a narrow band microstrip patch antenna operating at 2.45 GHz with a maximum gain of 5 dBi. The patch size is 28 mm×38 mm, which is associated with the center frequency of the antenna. The feedline width corresponding to a 50-ohm transmission line is 1.5 mm. Furthermore, the receiving antennas are monopole patch antennas with broad bandwidths. The antennas are fabricated using a substrate with dielectric constant, thickness, and loss tangent of 2.54, 0.54 mm, and 0.002, respectively.

C. EXPERIMENTAL SYSTEM SETUP
Figs. 9 and 10 depict the completed modules used for the experimental analysis.

III. EXPERIMENTS
The relationship between the permittivity of MUT and t d can be derived using (2) to (4).   where ε MUT , which is the main factor affecting the propagation time along the direction of delay line, denotes the relative permittivity of the substrate MUT, and A is a constant associ-  ated with the delay line structure to be derived experimentally. As (5) represents a simple proportional function between t d and √ ε MUT , a single test is sufficient to obtain the coefficient of A. Additionally, the actual distance between the tag and reader must be measured to determine the results accurately.
A. REFERENCE SIGNAL MEASUREMENT Fig. 11 depicts the unloaded test setup, wherein the transmitter antenna is placed directly below the receiver antenna without any tag. The reference signal V 0 (Fig. 1) is detected using this test, and the captured signal is illustrated in Fig. 12 with the x-axis and y-axis indicating the time and amplitude, respectively. Initially, the noise level is observed before the generation of the signal waveform. We recorded the first three peaks of the measured waveform and considered them as reference for the remaining tests. Fig. 13 depicts the air test setup, wherein the tag is placed along with the transmitter and receiver antennas. A test was performed to capture the received signal (V r + V 0 ) in the air. V 0 indicates the same signal observed in the unloaded test and  V r is the reflected signal from the tag. Fig. 14 illustrates the received signal observed by the oscilloscope. The waveform in Fig. 14 does not differ significantly from that observed in Fig. 12. This is because V r is close to V 0 and the two components nearly overlap with each other. The V 0 component in the air test is identical to that of V 0 obtained from the unloaded test without the tag. To obtain restored V r , we subtracted the waveform of V 0 in Fig. 12 from that of Fig. 14. As indicated in Fig. 15, the shapes of the first three peaks and first valley of the restored V r are identical to those of V 0 . Equation (1) calculates the time delay between the restored V r and V 0 , denoted by t measure .

B. AIR TEST AND EQUATION DERIVATION
The value of t measure was found to be 3.818 ns. Owing to the use of RFID technology in signal transmission, the distance s between the reader and tag is limited, the value of s was calculated to be 0.15 m. Furthermore, the travelling time of the signal in air can be determined using (6).
Considering the time spent in the RFID tag, the delay time is obtained using t d = t measure − 2t 1 = 2.818 ns. As the relative permittivity of air is ε air = 1, the value of A in (5) is obtained as 2.818. Thus, the relationship between the delay time in the tag and the MUT relative permittivity is obtained using (7).
where the unit of t d is seconds. Equation (7) is the final derived relationship used in the subsequent validation test.

C. VALIDATION TEST WITH WATER
We tested the permittivity of water at 25 • C using the derived relationship in (7). Fig. 16 depicts the experimental setup in water, wherein substantial loss is observed in the water environment owing to the tag being dipped in water. The thickness of the container is sufficiently thin to be ignored. The experiment was conducted identical to the air test. The result obtained is depicted in Fig. 17. Based on the simulation results, a time delay of few seconds is expected. As the two components in the received waveform do not overlap, a subtraction is not required. The three vertical lines in Fig. 17 indicate the three small peaks where V 0 exhibits the noise. The pattern of these three peaks with an oscillation after the first peak is identical to the one observed in Fig. 14. This pattern indicates the waveform reflected by the tag (V r ) as it cannot originate elsewhere. The time delay t measure between V 0 and V r is 31.12 ns. Thus, the relationship derived in (4) and (5) is used to calculate the permittivity of water.
In (1), t 1 denotes the time required for the signal to travel through the path between the reader and tag. The distance s of the path is 0.15 m, comprising the distance values of water S w = 0.1 m and air S a = 0.05 m, with the water relative permittivity of ε rw . Therefore, the delay time is measured VOLUME 9, 2021  using (8).
The relative permittivity of water is determined as ε rw = 78.1, considering t measure = 31.12 ns.
The air and water are applied as the actual operating environment in this measurement, and the error exists primarily in the measurement of the distance between the reader and tag.

IV. CONCLUSION
We designed a simplified chipless RFID permittivity sensor system comprising a reader and tag. The system was tested in air and water environments, and the corresponding signal processing procedures were determined in a stepwise manner. Furthermore, the experimental test was performed for individual components considering detailed circuit structures. We derived an equation to convert the measured signal into the relative permittivity value of the MUT. The system was calibrated to obtain a reference signal based on an unloaded test, and the measurement equations were determined through an air test. Subsequently, we conducted a validation test to obtain the value of water permittivity sensing. The water test was used to simulate the environment where the proposed system can be potentially applied, and the equations derived from the test were validated. The measured relative permittivity of water was 78.1. Considering the errors induced in the measurement, the result agreed with the actual value of relative permittivity of water. Thus, the proposed permittivity sensor system can be used to measure the permittivity of other materials.
This sensor can be applied in various environments or machines to test the permittivity of the liquid and evaluate the water quality. In the future, we intend to design a more compact structure to access unreachable target materials. Additionally, further investigations are necessary to obtain measurement methods with enhanced precision, better waterproof performance, higher working frequency, and advanced processing technology.