Structural health monitoring (SHM) has become a critical requirement for ensuring the safety and reliability of modern infrastructure systems. Compared with conventional nondestructive evaluation (NDE) techniques, SHM enables continuous, distributed monitoring of structural conditions, which is essential for long-term performance assessment of bridges, pipelines, railways, and industrial storage systems. This shift toward real-time condition monitoring has enabled predictive maintenance strategies that reduce both operational risk and lifecycle cost [1]. In high-risk industries such as oil and gas transportation, early detection of structural degradation is essential to prevent catastrophic failures. Defects such as micro-cracks, corrosion, and material fatigue gradually evolve and often remain undetected until structural integrity is severely compromised. In particular, dielectric-based structures used in insulation, coatings, and composite layers are highly vulnerable to crack propagation, which can significantly degrade system performance [2].

Although conventional NDE techniques such as ultrasonic inspection, eddy current testing, and thermography provide accurate defect detection, they are fundamentally limited by their requirement for direct access, high operational cost, and inability to support continuous monitoring over large or inaccessible areas [3]. These limitations motivate the development of autonomous sensing systems capable of wireless, scalable, and long-term operation.

Wireless sensing approaches based on radio-frequency identification (RFID) have emerged as promising alternatives. In particular, chipless RFID systems eliminate the need for integrated circuits by encoding information through electromagnetic resonant signatures. This enables fully passive sensing platforms that are low-cost, lightweight, and suitable for large-area deployment.

However, existing chipless RFID sensing approaches still exhibit critical limitations when applied to structural crack monitoring. Most reported designs rely on single or dual resonant features, which significantly restricts the amount of extractable sensing information. As a result, these systems typically provide only binary or low-resolution defect indication rather than true quantitative characterization.

Furthermore, commonly used resonator types such as dipoles, spirals, and patch-based structures suffer from inherent constraints including limited quality factor stability, polarization dependence, and weak sensitivity to localized dielectric perturbations. Environmental interference, fabrication tolerances, and reader alignment issues further degrade their reliability in practical SHM environments [4]-[7].

Recent advances have attempted to improve chipless RFID performance through multi-bit spectral encoding and hybrid amplitude-phase techniques. While these methods increase data capacity, they introduce additional system complexity, require precise phase coherence, and remain sensitive to multipath and environmental distortions, limiting their deployment in real-world structural monitoring scenarios [8][9].

More recently, multi-parameter RFID sensors have been reported for monitoring strain, displacement, and surface defects. Although these approaches demonstrate improved sensing capability, they are generally optimized for metallic substrates and exhibit reduced performance on dielectric materials due to weak field confinement and reduced interaction with localized defects [10]-[15].

To overcome these limitations, this work introduces a wideband, multiresonance, chipless RFID system specifically designed for quantitative crack-defect estimation in dielectric structures. Unlike conventional single-resonance or low-bit encoding tags, the proposed approach employs a compact LV-shaped resonator architecture that supports multiple stable, independent resonant modes across a wide frequency band. These resonances provide enhanced degrees of freedom for capturing subtle variations in crack geometry.

The key innovation of the proposed system lies in its ability to transform crack-induced dielectric perturbations into multi-resonance spectral signatures, enabling not only detection but also estimation of crack width and orientation with improved resolution. In addition, a wideband antipodal Vivaldi antenna is integrated as the interrogation unit, enabling efficient excitation and robust capture of resonance shifts under practical wireless conditions.

Unlike existing chipless RFID SHM systems, the proposed approach enables high-resolution, passive, and scalable crack characterization in dielectric structures, making it suitable for long-term monitoring of critical infrastructure.

The proposed sensing architecture is based on a wideband multiresonance chipless RFID tag that encodes structural information directly into its frequency-domain scattering signature, as illustrated in Figure 1. Unlike conventional RFID identification schemes that rely on single-resonance or binary spectral encoding, the proposed system exploits multiple independently excited resonant modes to significantly expand the sensing dimensionality for defect characterization. In this framework, the sensor tag operates as a passive electromagnetic transducer, where its scattering response is governed by the interaction between incident electromagnetic waves and the local dielectric environment of the monitored polyamide structure. Any variation in structural integrity, such as crack formation or propagation, perturbs the effective permittivity distribution around the tag, resulting in measurable shifts in multiple resonant frequencies.

The interrogation mechanism is based on a wideband radar-inspired excitation signal transmitted from the reader unit, enabling simultaneous excitation of all resonant modes within the tag. To enhance detection robustness and suppress direct coupling between transmitter and receiver, a cross-polarized configuration is employed. The transmitting antenna radiates a vertically polarized field, while the receiving antenna is oriented orthogonally to capture only the depolarized scattered response, thereby improving signal to clutter isolation in complex environments [16][17]. Upon excitation, the multiresonance tag exhibits a frequency-selective scattering behavior in which each resonant mode acts as an independent sensing channel. These resonances are not only determined by the geometric configuration of the tag but are also highly sensitive to localized dielectric perturbations induced by structural defects in the underlying material. This enables the conversion of spatial crack variations into a high-dimensional spectral signature. The received backscattered signal is processed in the frequency domain to extract the multiresonant signature of the tag. Changes in resonance frequency spacing, amplitude variation, and spectral distortion collectively provide a richer feature set for quantitative defect estimation, rather than simple binary detection.

Compared to conventional chipless RFID sensing approaches that rely on limited resonance count or single-parameter sensitivity, the proposed multiresonance architecture introduces an enhanced sensing framework capable of capturing higher-order electromagnetic interactions between the tag and the damaged dielectric medium. This results in improved resolution and increased robustness for structural health monitoring applications.

Figure 1 illustrates the operating principle of the proposed chipless RFID sensing system, including wideband interrogation, cross-polarized transmission/reception, and the depolarized backscattered response from the sensor tag.

Figure 1. Principle of operation of Chipless RFID sensor.

The proposed sensing element is a wideband multiresonance chipless RFID tag based on an integrated L-shaped and V-shaped depolarizing resonator configuration. The LV geometry is specifically engineered to support multiple orthogonal current paths and controlled electromagnetic coupling mechanisms, resulting in enhanced spectral richness for dielectric defect estimation in structural health monitoring applications. The geometrical layout and key design parameters are illustrated in Figure 2. Unlike conventional single-resonance chipless RFID tags, the proposed structure is not designed solely for frequency allocation but for multi-mode electromagnetic scattering generation, where each resonant segment contributes independently to the overall spectral signature. This enables higher information density and improved sensitivity to small variations in the surrounding dielectric environment.

The tag is fabricated on a Rogers RO3035 lossy substrate with dielectric constant ε r = 3.5 and loss tangent δ = 0.0023 . The top metallization layer consists of the LV-shaped copper resonator, while the bottom layer forms a continuous copper ground plane with a thickness of 0.018 mm. The monitored polyamide structure is positioned beneath the ground plane and acts as a lossy sensing medium, enabling direct electromagnetic interaction with the resonant field distribution. The substrate thickness is fixed at h = 1.52 mm, and both substrate and ground plane share identical lateral dimensions of 35 × 35 mm. The sensing medium (polyamide structure) is extended to 60 × 60 mm to ensure sufficient field penetration and to minimize edge reflection effects, thereby improving the reliability of dielectric interaction modeling.

The design methodology follows a sensing-driven resonator synthesis approach, where classical microstrip antenna theory is adapted not for radiation optimization but for maximizing dielectric perturbation sensitivity and multiresonance separation. The fundamental electromagnetic behavior is first related to resonant wavelength as:

where C is the velocity of light and f is the resonance frequency. The patch width is then defined using Equation (1.1).

w is the width of the patch, and the relative permittivity of the substrate is ε r . The effective electromagnetic behavior of the resonator is governed by the interaction between the physical geometry and the surrounding dielectric environment, which is expressed through

where C the speed of light in a vacuum L e f f is the effective length of the patch, and ε r e f f is the substrate’s effective relative permittivity L e f f takes into account both physical length and fringing fields at the patch’s tips, resulting in:

L ′ is the length of the patch which is equal to the effective length ( L e f f + 2 Δ L ), whereas Δ L is the extended length, Δ L can be further elaborated as

where w is the patch’s width and h is the thickness of the substrate underneath the patch. In both (1.3) and (1.5), ε r e f f is related to the dielectric constant ε r , w and h of the substrate, and may therefore be represented as:

The Equations (1.4) to (1.5) describe the relationships between the resonance of a V-shaped patch and several parameters, including the patch’s length and width, substrate thickness, and substrate permittivity. These estimate equations are valid for a V-shape patch constructed on a Roger substrate. Substrate length and width are calculated by using the following formulas, which are given below.

Whereas Lg & Wg are the length and width of the substrate, and h is given by

Figure 2 illustrates the configuration of the LV-shaped chipless RFID sensor and identifies the principal geometrical parameters used to define the resonant structure.

Figure 2. Configuration and geometrical characteristics of a chipless RFID sensor.

The detailed geometrical parameters of the proposed structure are provided in Table 1, which defines the dimensional variations of each resonant arm responsible for multiresonance generation.

Table 1. Dimension of Multiresonance Chipless RFID sensor.

This section established the principle of multiresonance generation for dielectric crack defect estimation using microstrip resonators [16]. Building on this concept, the present study employs a cross-polarized interrogation framework, where the chipless RFID tag operates as a depolarizing electromagnetic scatterer [16][17]. This enables robust separation of the transmitted and backscattered signals, improving detection reliability in practical structural health monitoring environments.

The proposed design is inspired by the concept of radio-frequency encoding particles [18], where multiple resonant scattering elements are engineered to encode information in the frequency domain. In this work, the concept is extended by integrating L-shaped and V-shaped resonant geometries into a unified LV-shaped architecture to achieve enhanced multiresonance behavior and improved dielectric sensitivity.

The final configuration of the proposed chipless RFID sensor tag is illustrated in Figure 3. The structure consists of a combination of diagonal dipole-based resonators arranged in L and V geometries with varying physical lengths. The V-shaped section is formed by two orthogonal sets of dipole arms: one set al.igned diagonally and the other oriented orthogonally, resulting in controlled current redistribution and strong depolarization effects [16][17]. Two additional L-shaped resonators are positioned adjacent to the V-shaped structure to further enhance multiresonance generation and spectral diversity.

From an electromagnetic perspective, this hybrid arrangement induces multiple current paths and coupling mechanisms within a compact footprint, leading to the formation of distinct resonant modes across the operating bandwidth. These resonances collectively form a wideband multiresonance spectral signature, which is highly sensitive to variations in the surrounding dielectric environment [18].

Each resonant element is etched on a Rogers RO-3035 substrate with a relative permittivity of 3.5 and a thickness of 1.52 mm. The total footprint of the sensor tag is maintained at 35 mm × 35 mm to ensure compact integration while preserving sufficient field confinement for stable resonance behavior. The monitored polyamide structure lies beneath the ground plane and interacts with the fringing electromagnetic fields, enabling defect-induced perturbations in the scattering response [19].

The proposed LV configuration is specifically designed to enhance mode diversity, polarization conversion efficiency, and resonance separation, which are critical factors for improving crack defect estimation accuracy [16]-[18]. Unlike conventional single-geometry chipless RFID tags, the hybrid LV structure introduces controlled asymmetry and orthogonal coupling paths, resulting in richer spectral encoding capability.

Figure 3 illustrates the full-wave simulation setup used to evaluate the multiresonance RCS response of the proposed chipless RFID sensor under plane-wave excitation.

The proposed chipless RFID sensing system is illustrated using a conceptual interrogation framework operating over the 2 - 6 GHz frequency range. The framework consists of two broadband antipodal Vivaldi antennas arranged in a cross-polarized configuration, where one antenna transmits the incident electromagnetic wave and the second antenna receives the backscattered response from the sensing tag. This configuration is presented to demonstrate a practical implementation of the proposed sensing concept and to illustrate how the tag may be interrogated in a real measurement environment.

As shown in Figure 10, the sensing tag is positioned on the dielectric specimen containing the crack defect. Variations in crack width and orientation modify the electromagnetic interaction between the resonator and the surrounding dielectric material, resulting in changes in the resonance frequencies of the scattered RCS spectrum. These resonance-frequency variations constitute the sensing signature used for crack characterization.

It should be emphasized that the reader configuration shown in Figure 10 was not explicitly modelled or experimentally implemented in this work. All resonance characteristics, frequency shifts, sensitivity values, and crack-characterization results presented in this paper were obtained exclusively from full-wave electromagnetic simulations performed in CST Studio Suite. The reported spectra therefore represent simulated monostatic radar cross-section (RCS) responses of the proposed chipless RFID sensor rather than signals acquired from a physical interrogation system.

In a practical implementation, the backscattered response could be measured using broadband antennas and conventional RF instrumentation, followed by resonance extraction using standard peak-detection or curve-fitting techniques. The design and experimental validation of such a hardware interrogation system are beyond the scope of the present work and will be considered in future studies.

Figure 10. Chipless RFID sensor with UWB low scattering and antipodal Vivaldi antenna (AVA).

The reader antennas are arranged in an orthogonal configuration to optimize polarization diversity and improve the robustness of backscatter detection. Specifically, the ultra-wideband (UWB) low-scattering antipodal Vivaldi antenna is oriented vertically, while the antipodal Vivaldi antenna (AVA) is positioned horizontally. This cross-oriented arrangement enhances isolation between transmitted and received signals, thereby improving the accuracy of chipless RFID tag interrogation.

To compensate for the inherently low backscattered power of chipless RFID tags, high-gain antennas are employed at the reader side, with gain values ranging from 7 dBi to 10.2 dBi across the 2 - 6 GHz operating band. This ensures sufficient link budget for reliable detection of weak multiresonance responses from the sensor tag under practical conditions.

The simulated RCS spectra obtained from CST Studio Suite were analyzed to evaluate crack-induced resonance variations. The six resonant modes shown in Figure 11 and Figure 12 constitute the sensing features used for crack characterization. The spectral response exhibits six distinct resonance peaks corresponding to the designed LV-shaped resonator structure.

Figure 11. The multiresonance chipless RFID sensor tag attached to the crack with varying crack orientations and crack widths: (a) at; (b) at.

Figure 11(a) illustrates the simulated multiresonance spectral response of the proposed chipless RFID sensor for dielectric cracks oriented at 0˚ with different crack widths. As the crack width increases, the electromagnetic coupling between the LV-shaped resonator and the polyamide substrate is progressively modified due to the reduction in the effective dielectric loading within the sensing region. This perturbation results in measurable variations in the resonance spectrum, where several resonant modes exhibit frequency shifts with different sensitivities. The response demonstrates that the proposed sensor can distinguish different crack widths while maintaining a stable multiresonance signature throughout the operating bandwidth. The observed resonance variations indicate that crack width directly influences the electromagnetic field distribution around the resonator, thereby providing a reliable indicator for structural defect characterization.

Figure 11(b)presents the simulated resonance response for 45˚ crack orientation. Compared with the 0˚ orientation, the resonant signatures exhibit a different frequency-shift pattern because the crack intersects the localized electric-field distribution at a different angle. Consequently, the electromagnetic interaction between the resonator and the dielectric material changes, resulting in orientation-dependent resonance behavior. Although the overall multiresonance characteristics are preserved, the relative sensitivity of individual resonances differs from that observed in Figure 11(a), demonstrating that the sensor is capable of discriminating crack orientation in addition to crack width.

Figure 12(a) shows the simulated resonance spectra for 90˚ crack orientation under different crack widths. Similar to the previous cases, increasing crack width modifies the effective dielectric environment surrounding the resonator, producing systematic perturbations in the resonance frequencies. The resulting spectral response remains clearly distinguishable for each crack condition, confirming that the proposed multiresonance sensor maintains consistent sensing capability irrespective of crack orientation. The differences in resonance behavior relative to Figure 11(a) and Figure 11(b) further indicate that the orientation of the crack significantly influences the electromagnetic coupling mechanism and consequently the resonance signature.

Figure 12. The multiresonance chipless RFID sensor tag attached to the crack samples with varying crack orientations and crack widths: (a) at; (b) at.

Figure 12(b) illustrates the spectral response for 135˚ crack orientation. The resonance characteristics continue to exhibit orientation-dependent behavior, with several resonant modes showing distinguishable frequency variations as the crack width increases. These variations arise from changes in the distribution of the electric fields around the resonator caused by the altered crack geometry. The unique resonance signature associated with this orientation further demonstrates the capability of the proposed chipless RFID sensor to identify multiple crack configurations using a single passive sensing structure.

Overall, Figure 11 and Figure 12 demonstrate that the proposed multiresonance chipless RFID sensor produces distinct and repeatable resonance signatures for different combinations of crack width and orientation. Rather than relying on a single resonant frequency, the sensor exploits the collective response of multiple resonant modes, thereby improving defect discrimination and enhancing sensing reliability. Since each crack orientation produces a unique resonance pattern, the multiresonance response provides sufficient information to characterize both the severity and orientation of dielectric cracks. All resonance spectra presented in Figure 11 and Figure 12 were obtained from full-wave electromagnetic simulations using CST Studio Suite and represent the simulated backscattered response of the proposed interrogation framework. No experimental measurements, receiver-voltage acquisition, or hardware signal-processing implementation were performed in this study.

Table 3. Comparison on proposed work and previous studies related to chipless RFID sensor designs for crack detection.

Table 3 compares the proposed multiresonance chipless RFID sensor with representative prior sensor designs in terms of sensing method, measured parameter, operating frequency range, and reported advantages and limitations.

In comparison with existing state-of-the-art chipless RFID sensing approaches, the proposed multiresonance LV-shaped sensor demonstrates clear improvements in crack detection capability, operational robustness, and structural health monitoring functionality. Prior works such as [12] and [8] primarily focus on single-parameter sensing and are largely constrained to metallic surfaces, limiting their applicability in dielectric-based structural monitoring scenarios.

In contrast, the proposed design introduces a multiresonant LV-shaped architecture that enables simultaneous extraction of crack width and orientation through distinct frequency shift patterns. This significantly enhances the information-carrying capacity of the sensor compared to conventional single- or dual-resonance chipless RFID tags.

Compared to spiral resonator-based sensing techniques [5], which exhibit sensitivity to environmental variations such as humidity and suffer from limited resolution stability, the proposed system maintains consistent multiresonance behavior across a wide UWB operating band (2.3 - 5 GHz). This results in improved frequency stability, higher encoding reliability, and enhanced defect discrimination capability.

Furthermore, the integration of a UWB antipodal Vivaldi antenna improves radiation efficiency and bandwidth performance compared to narrowband reader architectures commonly reported in earlier studies. The proposed system also benefits from passive operation, low cost, and scalable deployment, while calibration-based signal processing effectively mitigates noise and false peak detection issues.

Overall, the proposed chipless RFID sensor overcomes key limitations of existing approaches, including substrate dependency, limited read range, and single-parameter sensing, thereby providing a more robust and adaptable solution for dielectric crack monitoring in structural health monitoring applications.

This study presented the design and analysis of a wideband multiresonance chipless RFID sensor for the detection and characterization of dielectric cracks in structural materials. The proposed LV-shaped resonator architecture demonstrates that key crack parameters, including width and orientation, can be effectively extracted through distinct multiresonance frequency shift patterns.

The results confirm that the developed sensing framework, combined with cross-polarization interrogation and an antipodal Vivaldi antenna-based reader, enables reliable detection of defect-induced variations in the frequency signature. The proposed system successfully characterizes crack orientations of 0˚, 45˚, 90˚, and 135˚, as well as crack widths ranging from 1 mm to 3 mm.

A depolarizing chipless RFID tag configuration, together with controlled near-field interrogation at a fixed reading distance of approximately 15 cm, ensures stable extraction of the multiresonance spectral response even in the presence of a large monitored dielectric structure. The use of cross-polarized measurement further enhances signal isolation and improves detection robustness.

Overall, the proposed LV-shaped multiresonance chipless RFID sensor provides a compact, passive, and high-resolution sensing solution with strong radar cross-section performance. Compared with conventional chipless RFID approaches, it offers improved feature richness, enhanced defect discrimination capability, and better suitability for practical structural health monitoring applications.

The presented approach demonstrates strong potential for future deployment in infrastructure monitoring systems, where low-cost, scalable, and wireless non-destructive evaluation techniques are required. Future work may focus on extending read range, improving multi-tag discrimination, and experimental validation in real-world environments.

The results presented in this study highlight the strong potential of the proposed chipless RFID-based sensing system for structural health monitoring of critical infrastructure, including railways, pipelines, and oil storage tanks. Despite these promising outcomes, several practical challenges must be addressed to enable large-scale field deployment and system integration.

A key limitation lies in improving the maximum reading range and overall detection accuracy. Future design optimization should focus on identifying an optimal tag topology that balances spectral encoding capacity, spatial resolution, and interrogation distance. Although miniaturization of chipless RFID tags remains an active research direction, it presents significant challenges due to inherent trade-offs between size reduction and resonance stability.

Further research is also required to overcome current technological constraints and to expand the application scope of passive multiresonance chipless RFID sensors. Future developments can be broadly categorized into three main directions.

1) Networking and standardization: development of tag arrays and controlled tag-tag coupling mechanisms for improved coverage, along with integration of ultra-wideband (UWB) techniques for high-capacity data encoding and robust interrogation.

2) Flexibility and ubiquity: integration with hybrid sensing platforms combining chipless RFID with embedded sensor technologies, implementation of automated impedance matching and digital RSSI processing, exploration of functional materials for analog memory behavior, and adaptation for wearable and biomedical sensing applications.

Overall, these directions indicate a pathway toward more intelligent, scalable, and multifunctional chipless RFID systems for next-generation structural health monitoring and sensing networks.

The authors would like to express their sincere appreciation to the School of Information and Communication Engineering, Northwestern Polytechnical University, for providing academic and research support during the development of this work. The authors also gratefully acknowledge the valuable technical discussions and suggestions received during the preparation of this manuscript.

Conceptualization, M.Z.; methodology, M.Z.; software and simulation, M.Z.; formal analysis, M.Z.; investigation, M.Z.; data curation, M.Z.; visualization, M.Z.; writing—original draft preparation, M.Z.; writing—review and editing, F.Z.K., M.K.S. and A.G.A.; supervision and technical guidance, F.Z.K., M.K.S. and A.G.A. All authors have read and agreed to the published version of the manuscript.