Design and fabrication of a TE 50 mode SIW-fed Yagi antenna array at 5.8 GHz

1 Introduction

Modern wireless communication systems increasingly demand antennas with advanced performance characteristics, such as high directivity, significant gain, wide bandwidth, and ease of integration. In this context, Yagi-Uda antennas—renowned for their high directivity and gain—continue to play a pivotal role in various applications. Simultaneously, Substrate Integrated Waveguide (SIW) technology has emerged as a promising solution to address challenges related to the miniaturization and integration of microwave components (Omrani et al., 2024; Sun et al., 2017; Sun et al., 2022a). Initially developed by Deslandes and Wu (2001) and further expanded by Prof. Xu and Hong (2011), SIW technology represents a significant advancement in waveguide design. It aims to replace conventional metallic waveguides while preserving their electromagnetic advantages. SIW structures are particularly attractive due to their low profile, reduced cost, lightweight nature, and ease of integration (Xu and Hong, 2011; Guan et al., 2014; Li and Dou, 2015). These planar waveguides emulate the behavior of traditional metallic waveguides by utilizing rows of metallized via holes through the substrate that connect the top and bottom conductive planes Cai et al. (2015).

While conventional substrate integrated waveguides (SIWs) predominantly operate in the fundamental TE 10 mode, scaling these structures to large-scale antenna arrays introduces severe manufacturing constraints. Specifically, standard printed circuit board (PCB) and photolithography processes impose strict physical tolerances, such as a minimum metallization/slot etch resolution of 100 μ m and a drill positioning accuracy of ± 25 μ m for the via diameter ( d ) and inter-via pitch ( p ). At higher frequencies, dense via arrays with small diameters ( d < 0.2 λ 0 ) introduce localized phase errors, radiation leakage, and drill overlap defects Omrani et al. (2024); Verma et al. (2023); Sharma and Budhiraja (2024). To mitigate these quantitative fabrication limitations and reduce via density, recent research has leveraged higher-order SIW modes to achieve wider transverse apertures without requiring dense internal metallic via walls.

The exploration and exploitation of higher-order modes in substrate integrated waveguides (SIWs) have progressed significantly to overcome the aperture constraints of the fundamental mode. Early foundational analyses established the propagation behavior of higher-order modes, such as the even TE 20 mode Deslandes and Wu (2003) and the TE 30 mode Cassivi et al. (2002). Subsequent developments expanded this paradigm to higher modes, including the TE 40 mode Yan et al. (2004) and modal field analysis up to the TE 50 mode Xu and Wu (2005). Each mode offers distinct spatial field properties tailored to specific antenna applications. In particular, the TE 50 mode provides substantial advantages for high-gain planar structures. Its transverse electric field distribution—characterized by five distinct field maxima across the cavity width—enables a broader radiating aperture and more uniform spatial excitation. Consequently, operating in the TE 50 mode allows the design of wide aperture antennas with reduced via-wall density, effectively simplifying fabrication tolerances and reducing manufacturing complexity without sacrificing radiation efficiency.

Although conventional substrate-integrated waveguide (SIW) cavity structures are primarily designed for broadside radiation Cheng et al. (2010); Luo et al. (2008), alternative configurations leverage SIW cavity modes—such as the TE 30 mode—to excite end-fire antenna topologies Sun et al. (2017), Sun et al. (2022a). Building upon this concept, the main novelty of the proposed design lies in the use of the higher-order TE 50 mode in an SIW feeding structure to excite an expanded five-element Yagi antenna array with five directors at 5.8 GHz. This approach extends previous TE 30-mode SIW-fed configurations to a higher-order modal excitation, enabling a larger antenna array within a compact planar structure.

Notably, while our preliminary conference paper Omrani et al. (2024) introduced the initial conceptual idea and basic bandwidth simulations of a TE 50-mode SIW feed, this journal work provides a comprehensive structural, theoretical, and experimental investigation. Specifically, this manuscript significantly expands upon the preliminary study by introducing: (i) a detailed physical and equivalent circuit analysis of the dual-slot mode conversion process, (ii) extensive parametric evaluations governing mode purity, (iii) the physical fabrication and experimental measurement of the antenna prototype in an anechoic chamber (validating return loss, 2D radiation patterns, and peak gain exceeding 10 dBi), and (iv) a quantitative performance benchmark against state-of-the-art literature.

To overcome the inherent technical challenges associated with the efficient excitation of this higher-order TE 50 mode within the overmoded SIW cavity (W SIW = 127 mm, ε r = 2.2), an original feeding mechanism is proposed. The architecture employs a symmetric two-branch T-divider on the upper substrate to achieve balanced power splitting and equal phase excitation. This microstrip network feeds two strategically positioned longitudinal slots on the middle metal layer. Rather than directly dividing the signal five ways, these slots act as discrete spatial excitation sources that excite the target electric field antinodes. Upon coupling into the SIW cavity, spatial modal synthesis and resonant boundary conditions enforce constructive interference at 5.8 GHz. This mechanism selectively transforms the two slot feeds into the required five-peak field profile (+, -, +, -, +) while suppressing parasitic lower-order modes, as verified by the E z-field distribution in Figure 1.

Although various Substrate Integrated Waveguide (SIW) feeding techniques have been extensively reported in the literature to excite antenna arrays, most conventional designs primarily rely on the fundamental TE 10 mode. Although effective, TE 10 mode feeding networks often suffer from structural bulkiness or complex corporate-feed distribution when scaling up to directive multi-element arrays such as Yagi-Uda antennas.

To overcome these limitations, this article proposes a novel and compact feeding mechanism operating entirely in the higher-order TE 50 mode at 5.8 GHz. The distinct novelty of this work lies in the unique integration of a microstrip T-type power divider combined with strategically positioned dual slots for efficient mode conversion. Unlike existing SIW-fed architectures, this specific configuration achieves a highly directive radiation pattern (with’measured realized gain of 10 dB at 5.8 GHz) while drastically simplifying the feeding routing and minimizing the overall footprint of the antenna array. Furthermore, the fabrication tolerances are significantly improved, offering a cost-effective and structurally optimized solution compared to the state-of-the-art designs.

To overcome the efficiency limitations of wide-aperture arrays, an oversized Substrate Integrated Waveguide (SIW) feeding network operating in the higher-order TE 50 mode is proposed for driving the 5.8 GHz Yagi-Uda antenna array. To quantitatively justify the selection of the TE 50 mode over lower-order modes (TE 10 and TE 30) for driving the broad-aperture array, a comprehensive parametric trade-off study was conducted at 5.8 GHz using analytical EM formulations and full-wave HFSS simulations, as summarized in Table 1. While a conventional fundamental TE 10 guide offers a broader fractional bandwidth, illuminating a 127 mm wide Yagi-Uda array would require a complex multi-stage power divider network, introducing additional dielectric/metallic losses, phase errors, and layout complexity. Conversely, utilizing the oversized SIW cavity in the higher-order TE 50 mode naturally establishes five spatially distributed electric field maxima across the 127 mm width. This provides a direct, highly efficient 5-point excitation for the radiating elements while reducing the conductor attenuation constant down to 0.78 dB/m. Although the higher-order mode exhibits a narrower fractional bandwidth ( 3,4 % ), it fully covers the required 5.8 GHz ISM band ( 5.66 - 5.94 GHz ) while ensuring high fabrication tolerance and ease of single-layer PCB integration.

TABLE 1. Quantitative trade-off analysis of modal feeds at 5.8 GHz (calculated analytically and simulated via HFSS).

The remainder of the paper is organized as follows: a detailed description of the design and fabrication of the proposed antenna is provided in Section 2. Then, Section 3 compares simulated and experimental results. Finally, concluding remarks are given in Section 4.

2 Description of the antenna design

The proposed structure comprises two main components: a five-element Yagi-Uda antenna array and an excitation structure operating in the higher-order TE₅₀ mode, which primarily generates five signals with equal power levels.

Figure 2 shows the geometric structure of the proposed Yagi-Uda antenna array. The main body of the antenna is integrated with an excitation structure operating in TE₅₀ mode.

2.1 Elementary Yagi-Uda antenna

For this design, Figure 3 was implemented on a dielectric substrate with the following parameters: Rogers RT/duroid 5880™ substrate, relative permittivity ε r = 2.2, substrate thickness h = 1.575 mm, and loss tangent tan ⁡ δ = 0.0018. The conducting material is copper with a metallization thickness of 35 μm (see Figure 3). The first director is placed at a distance of D y 1 = 4 mm from the feed dipole, while a uniform spacing of 2.5 mm separates the subsequent directors.

2.2 SIW power divider operating in TE 50 mode

The SIW power divider consists of two parts: the mode transition structure and the feed network implemented using a T-type microstrip power divider (Figure 4).

The T-type power divider plays a crucial role in our system by feeding the different branches of the higher-order mode divider in a balanced manner. By ensuring a uniform power distribution and minimizing reflection losses, it contributes to optimizing the overall performance of the device. The T-type microstrip power divider, with a line width w 1, feeds the mode transition structure (see Figure 5). The line width w 1, a critical parameter, was carefully adjusted to achieve a 50 Ω input impedance, enabling optimal power transmission. The equal length L 2 of the two branches of the T-divider ensures an equitable power split between the two outputs, with output signals exhibiting identical characteristics.

The SIW power divider consists of two components: the mode transition structure and the power input implemented using a T-type microstrip power divider, as illustrated in Figure 4. The spacing between the two slots is set to one-fifth of the width of the substrate-integrated waveguide (SIW). This configuration ensures a uniform power distribution among the five signals generated in the higher-order TE₅₀ mode (see Figure 4). The optimization of the position and the width parameters ( d via ) of the impedance-matched metallic vias guarantees maximum performance of the electric field in the higher-order mode.

To validate the physical operating principle and mode conversion mechanism from the quasi-TEM microstrip line to the higher-order TE 50 mode, the simulated vertical electric field distribution ( E z ) inside the SIW cavity at 5.8 GHz is depicted in Figure 1. The dual coupling slots are strategically positioned at the phase centers corresponding to the electric field antinodes of the target TE 50 mode. This spatial alignment enforces an equal power division with the required phase condition across the waveguide aperture, effectively suppressing lower-order parasitic modes (such as TE 10 and TE 30). As observed in Figure 1, exactly five distinct field intensity peaks with alternating phase polarities (+, -, +, -, +) are established along the SIW width W SIW, confirming high excitation purity, power balance, and strong dominance of the desired TE 50 mode. This field distribution can be explained by the transverse field variation of the T E 50 mode, which can be expressed by Equation 1 aswhere a eff is the effective SIW width. The positions of the five field maxima are given by Equation 2:

E y x ∝ sin 5 π x a eff , (1)

x n = 2 n − 1 a eff 10 , n = 1,2 , … , 5 . (2)

Therefore, the two slots do not directly generate five independent feeding points; instead, they excite the T E 50 modal distribution, whose five field maxima provide five effective excitation regions for the Yagi antenna array.

To further assess the modal behavior, the cutoff frequencies of the supported T E m 0 modes are considered. For an SIW with an effective width a eff, the cutoff frequency can be approximated by Equation 3 as

f c , m 0 = m c 2 a eff ε r . (3)

The cutoff-frequency analysis, together with the full-wave field distribution in Figure 1, supports the identification of the desired T E 50 mode at 5.8 GHz. In particular, the T E 10 and T E 30 modes exhibit one and three transverse field maxima, respectively, whereas the simulated field distribution exhibits the five alternating field maxima characteristic of the T E 50 mode.

2.3 Detailed design procedure and parameter roles

To ensure full design reproducibility, the geometrical parameters listed in Table 2 are classified according to their specific functional roles within the feed network and antenna elements:

• SIW Cavity and Transition: W siw defines the TE 50 mode cut-off frequency, d via and p specify the via-wall pitch and diameter, while W t and L t correspond to the total ground plane dimensions.
• T-Divider and Coupling Slots: W 11 and L 1 control the 50 Ω input impedance matching, L 2 sets the power division branch length, L s and W fs determine the dual coupling slot resonance, W btr sets the slot transformation width, and L 3 provides fine phase tuning.
• Yagi-Uda Array Elements: L y 1 to L y 6 represent the optimized lengths of the driven element, reflector, and directors, d y 1 and d y 2 denote the inter-element spacings, and W y p sets the director strip width.

TABLE 2. Geometrical parameters of the proposed antenna.

Complete design methodology follows a systematic four-step procedure, illustrated in the flowchart in Figure 5:

• Step 1 (SIW Cavity Sizing): Calculate W siw to support the TE 50 mode at the target center frequency using standard SIW equivalent width formulas.
• Step 2 (Feed Network Tuning): Optimize the T-divider arms ( L 1 , L 2 ) and dual slot geometry ( L s , W fs , W btr ) to achieve strong spatial field coupling and equal amplitude/phase excitation.
• Step 3 (Yagi Element Synthesizing): Determine initial lengths (L y 1–L y 6) and spacings ( d y 1 , d y 2 ) based on surface-wave launch conditions.
• Step 4 (Parametric Fine-Tuning): Conduct multi-objective parametric sweeps in HFSS to maximize gain and front-to-back ratio while maintaining S 11 < − 10 dB.

3 Parametric study

To ensure the optimal performance, stability, and impedance matching of the proposed antenna system, a comprehensive parametric analysis was conducted using ANSYS HFSS. Because structural dimensions heavily influence the resonant frequency, reflection coefficient, and power distribution, key parameters were systematically varied.

3.1 Feeding network and tuning stubs

The performance of the feeding network is crucial for achieving balanced excitation. Figure 6 illustrates the simulated S-parameters of the T-junction power divider, including the reflection coefficient S 11 and the transmission coefficients (S 21 and S 31). As shown, the input reflection coefficient S 11 remains below − 25 dB at the operating frequency of 5.8 GHz, indicating excellent impedance matching. Furthermore, the transmission parameters S 21 and S 31 achieve equal power splitting with values close to − 3.5 dB across the band. The tuning stubs integrated at the ends of the power divider branches compensate for reactive discontinuities, ensuring proper impedance matching and maintaining balanced power division between the output ports.

3.2 Antenna structural parameters

In addition to the feeding network, the effects of critical antenna dimensions on the reflection coefficient were investigated:

• Feed Slot Width ( W f s ): The variation of the feed slot width directly impacts the electromagnetic coupling between the microstrip line and the radiating elements. Figure 7 demonstrates its effect on the S 11 reflection coefficient, showing optimal resonance tuning around the target frequency.
• Via Diameter ( d vias ): The diameter of the substrate-integrated vias dictates the boundary conditions of the propagation structure. Figure 8 illustrates the stability of the reflection coefficient under varying via diameters.
• Slot Length ( L s ): The length of the coupling slot controls the electrical length and resonant path. As depicted in Figure 9, optimizing L s ensures a deep impedance bandwidth centered at the desired operating frequency.

4 Implementation and experimental validation

The proposed antenna is designed on a Rogers 5880 dielectric substrate with a relative permittivity of 2.2, a thickness of 1.575 mm, and a dielectric loss tangent of 0.0018. The detailed geometrical parameters of the antenna are summarized in Table 2. Figure 10 shows the photograph of the fabricated antenna prototype. Fig. Figure 11 presents the simulated and measured reflection coefficient ( S 11 ), which remains below −10 dB over the frequency range from 5.7 GHz to 5.9 GHz, achieving a relative bandwidth of 3.4%. The maximum measured gain is about 10 dBi in the entire operating band. The Measured and simulated radiation patterns of the antenna in the H-plane and E-plane at 5.8 GHz are shown in Figure 12 respectively. It can be seen that the antenna exhibits stable radiation patterns with a broad main lobe and low side lobes in both planes. The measured results are in good agreement with the simulated ones, with slight deviations attributed to fabrication tolerances and measurement uncertainties. The 3 dB bandwidths are approximately 85° in the E-plane and 90° in the H-plane at 5.8 GHz. To quantitatively substantiate the advantages of the proposed TE 50-mode topology, Table 3 provides a comprehensive performance benchmark against recent state-of-the-art SIW-fed planar arrays and higher-order mode antennas Sun et al. (2022b); Yin et al. (2023); Verma et al. (2023); Sharma and Budhiraja (2024); Luo et al. (2026). Compared to conventional fundamental and lower higher-order mode designs (such as TE 30 and TM 30), the proposed design achieves a competitive peak gain of 10.2 dBi at 5.8 GHz within a single-layer substrate. More importantly, operating in the TE 50 mode expands the transverse electric field distribution, enabling effective aperture excitation with a significantly reduced total metallic via count. By eliminating the need for dense internal via walls typically required for modal isolation or multi-layer stacked configurations, the fabrication tolerance constraints (± 25 μ m drill accuracy) and production costs are substantially mitigated without compromising impedance bandwidth ( 3.4 % ) or radiation performance.

TABLE 3. Quantitative performance comparison with higher-order mode and SIW-fed planar arrays.

5 Conclusion

This study presented the design, fabrication, and characterization of a Yagi-Uda antenna array fed by a Substrate Integrated Waveguide (SIW) operating in the higher-order TE 50 mode at 5.8 GHz. By directly exciting the TE 50 mode via a dual-slot transition network, the design circumvents multi-stage TE 10 corporate feeding networks. The fabricated prototype achieves a peak realized gain of 10.2 dBi and a fractional bandwidth of 3.4%. This architecture provides a compact planar option for directive sub-6 GHz wireless and radar systems. Future work will investigate structural optimization for bandwidth extension and array scalability.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fanpr.2026.1932708/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Sahar Saleh, Atlantic Technological University, Ireland

Reviewed by: Ruslan Politanskyi, Chernivtsi University, Ukraine

Sathish M, Rajalakshmi Engineering College, India

Tuan-Anh Le Trong, Viettel Group, Vietnam