High-Gain Dual-Band Microstrip Antenna for 5G mmWave Applications: Design, Optimization, and Experimental Validation

Abstract

This study presents a novel dual-band microstrip antenna operating at 28/38 GHz, which is designed for fifth generation (5G) and next-generation communications. The objective was to create a high-gain, single-element solution that addresses millimeter-wave (mmWave) challenges, like attenuation and signal loss, offering a more efficient alternative to complex array antennas. The antenna was designed using Rogers RT/duroid 5880 as a substrate, and CST simulations were used to optimize the return loss, gain, and efficiency. Analytical methods and parametric analyses were used to further optimize the design. Additionally, an SMP connector was integrated into the simulated model using Antenna Magus software, followed by further refinement through additional parametric studies. The final compact antenna (33 × 27 × 1.6 mm³) demonstrates excellent performance with simplified fabrication. The antenna achieved bandwidths of 1.12 GHz at 28 GHz and 1.27 GHz at 38 GHz, with remarkably low return loss values of −53.04 dB and −83.65 dB, respectively. The gain values reached 7.82 dBi at 28 GHz and 8.98 dBi at 38 GHz—prototype measurements closely aligned with simulations, confirming reliability. This study introduces a high-performance, single-element antenna that is both simple and complex. The meticulous optimization process, including SMP connector variations, minimized the fabrication sensitivity and improved the overall performance, thereby marking a significant advancement in antenna design.

1. Introduction

The increasing diversity of internet applications has driven the exponential growth of wireless communication systems, the rapid expansion of machine-to-machine (M2M) communication, and the proliferation of connected devices. The emergence of data-intensive applications, such as augmented reality (AR), virtual reality (VR), and ultra-high-definition (UHD) video streaming, has further accelerated the demand for ultra-high data rates, reaching up to 10 Gbps, as a fundamental requirement for modern communication networks [1]. However, conventional sub-6 GHz frequency bands are becoming increasingly congested, thereby imposing significant limitations on network capacity and hindering further performance improvements [2].

To address this challenge, millimeter-wave (mmWave) technology has emerged as a promising solution, offering access to largely underutilized frequency bands with significantly wider bandwidths, thereby enhancing spectral efficiency and enabling high-speed data transmission [3]. Despite these advantages, mmWave signals suffer from inherent challenges, including high free-space path loss and increased susceptibility to atmospheric absorption, which degrade signal propagation [4]. Consequently, developing highly efficient antennas capable of maximizing the gain while maintaining a compact and cost-effective design is crucial for the widespread adoption of mmWave technology [5].

Various antenna models have been explored for mmWave communication, yet many existing designs fail to meet the stringent requirements of future wireless systems. Although traditional aperture antennas can provide directional radiation characteristics, they are often impractical because of their large size, complex structures, and integration challenges [6]. As a result, microstrip antennas have gained significant attention owing to their lightweight structure, compact form factor, low fabrication cost, and ease of integration with planar circuits, making them strong candidates for high-frequency applications [7].

One critical design parameter of a microstrip antenna is its operating frequency. Regulatory bodies worldwide have designated specific 5G mmWave frequency bands. The United States has allocated the 27.5–28.35 GHz and 37–40 GHz bands, while Japan has adopted the 27.5–28.28 GHz range. Similarly, Europe has designated the 24.25–27.5 GHz spectrum, and China is expected to commercialize the 24.25–27.5 GHz and 37–43.5 GHz bands in next-generation communication networks [8]. Among these bands, the 28- and 38-GHz bands have garnered significant attention owing to their potential for widespread deployment and relatively favorable propagation characteristics, making them a primary focus in antenna research [9,10].

In mmWave antenna research, significant efforts have been made to enhance the antenna gain to mitigate propagation losses and extend the coverage. Traditional single-element antennas generally struggle to achieve gain values exceeding 5 dBi, which is insufficient for practical applications [11]. As a result, multi-element configurations, such as phased arrays and multiple-input multiple-output (MIMO) antenna systems, have been widely investigated to improve the gain and directivity [12,13]. Although these array structures provide enhanced performance, they also introduce additional design complexities, increased power consumption, and higher fabrication costs [14,15]. Hence, striking a balance between high gain and design simplicity remains a key research challenge.

Early research efforts in mmWave antenna design typically began with single-element configurations, which were later expanded into array structures to enhance key performance parameters, such as the gain, bandwidth, and beamforming capability [16,17]. However, achieving a high-gain, dual-band single-element antenna that can outperform array configurations remains an open challenge.

This study aims to address this challenge by introducing a novel high-gain, dual-band microstrip antenna designed for operation at 28 GHz and 38 GHz. The proposed antenna uses advanced parametric optimization techniques and innovative material selection strategies to enhance its gain and efficiency while maintaining a compact structure. By integrating rigorous performance evaluations and experimental validations, this research contributes to the development of practical, high-performance mmWave antennas for next-generation wireless networks. The proposed design aims to bridge the gap between single-element and array configurations, offering a viable alternative for mmWave applications, in which size, cost, and efficiency are critical considerations.

1.1. Research Gaps in Previous Research

Despite extensive research on mmWave antennas, single-element microstrip designs have often been overlooked in favor of array configurations, which are typically preferred for their higher gain and improved return loss. However, this study challenges the prevailing notion that single-element antennas are inherently inferior by demonstrating that a carefully optimized design can achieve performance metrics comparable to and, in some cases, exceeding those of array-based solutions. The proposed antenna delivers superior gain and achieves lower return loss (S₁₁), making it a competitive alternative to mmWave antennas.

A major gap in the existing literature is the scarcity of high-performance single-element microstrip antennas tailored for commercial mmWave frequency bands, such as 28 GHz and 38 GHz, which are crucial for 5G and beyond. Prior studies have predominantly focused on multi-element arrays to meet the stringent performance demands of these bands, often overlooking the potential of well-engineered single-element solutions. This work fills this void by demonstrating that a single-element design can provide high gain and low return loss while maintaining a compact and cost-effective structure when meticulously optimized.

In addition, this study highlights a frequently neglected factor in mmWave antenna research: the influence of the connectors on the overall performance. The integration of an SMP connector and its impact on key parameters, such as return loss and gain, were systematically investigated. Unlike many studies that either assume negligible connector effects or fail to discuss them altogether, this research provides a detailed analysis, offering valuable insights for practical implementations.

Furthermore, an extensive parametric analysis was conducted to enhance the robustness of the proposed design against manufacturing variations. By evaluating the sensitivity of each patch edge to the fabrication tolerances, the antenna was fine-tuned to ensure consistent performance. This approach significantly mitigates manufacturing-induced deviations, which is a challenge commonly faced in high-frequency antenna fabrication. The resulting design achieves remarkably low return loss values at the targeted frequencies, reinforcing its superiority over existing single-element antennas in the literature.

1.2. Contributions of This Paper

This study introduces a novel high-gain, dual-band microstrip antenna designed for operation at 28 GHz and 38 GHz, offering a compact, cost-effective, and high-performance alternative to conventional array configurations. The proposed design challenges the prevailing assumption that single-element antennas are inherently limited in performance by demonstrating that the gain and return loss values rival and, in some cases, exceed those of multi-element array systems. These findings provide new possibilities for efficient antenna solutions in 5G and beyond. The key contributions of this work are as follows:

• The proposed antenna achieves industry-leading return loss values of −53.04 dB at 28 GHz and −83.65 dB at 38 GHz, thereby setting a new benchmark for single-element microstrip antennas. Additionally, the measured gain values of 7.82 dBi at 28 GHz and 8.98 dBi at 38 GHz surpass those of comparable designs in the literature, demonstrating the suitability of the proposed design for high-performance mmWave applications.
• This study systematically examines the impact of an SMP connector on mmWave antenna performance, providing critical insights into its effects on the return loss and gain. Unlike most studies that assume negligible connector influence, this research highlights the significance of the influence and offers practical guidelines for real-world antenna integration.
• A rigorous optimization framework that combines analytical methods with parametric analyses was used to refine the antenna geometry. The influence of each patch edge on the S₁₁ parameter was carefully evaluated to ensure minimal sensitivity to the fabrication tolerances. This approach enhances antenna reliability and manufacturability, which are critical factors for large-scale deployment.
• With dimensions of only 33 × 27 × 1.6 mm³, the proposed design minimizes material usage and manufacturing complexity while maintaining superior performance. The compact structure and simplified fabrication requirements make it highly viable for commercial applications where size, cost, and efficiency are paramount.

2. Antenna Design Stages

The proposed dual-band microstrip antenna was developed systematically to ensure optimal performance at 28 GHz and 38 GHz for mmWave applications. The design methodology integrates precise dielectric material selection, rigorous analytical modeling, and parametric optimization, all of which are critical for achieving high gain, low return loss, and enhanced radiation efficiency in next-generation communication systems.

2.1. Antenna Design and Configuration

The dielectric substrate selection plays a pivotal role in determining the microstrip antenna impedance characteristics, bandwidth, and radiation efficiency, particularly at mmWave frequencies. To minimize signal attenuation [18] while maintaining structural stability, this study employs Rogers RT/duroid 5880, a well-established high-frequency substrate with a dielectric constant (ε_r) of 2.2 and a low loss tangent (tan δ = 0.0009). The substrate thickness (h) was set to 1.6 mm to ensure a balance between mechanical robustness and optimal electromagnetic performance.

The Rogers RT/duroid 5880 is particularly advantageous for millimeter wave applications, where excessive propagation losses can significantly degrade antenna efficiency. The low-loss properties of this material make it highly suitable for 5G and other systems, in which maintaining signal integrity and minimizing power dissipation are paramount.

Once the dielectric material was selected, the physical dimensions of the radiating patch, specifically its width (W) and length (L), were determined using well-established microstrip antenna design principles [19]. Achieving dual-band operation in single-element microstrip antennas requires approaches beyond classical mathematical models. The most commonly used techniques in the literature include slot configurations, the incorporation of notches in the patch geometry to achieve similar effects, the use of multiple resonant structures on the patch, the integration of short-circuit pins, and adjustments to the feed line configuration.

For the designed antenna to operate at dual-band frequencies of 28/38 GHz, the effect of notches on the patch is analyzed using slot configurations. The resonant frequencies (f_mn) for a classic rectangular microstrip antenna are given in Equation (1) [19,20,21].

$${\mathrm{f}}_{\mathrm{m}\mathrm{n}}={\displaystyle \frac{\mathrm{c}}{2\sqrt{{\epsilon }_{\mathrm{e}\mathrm{f}\mathrm{f}}}}}\sqrt{{\left({\displaystyle \frac{\mathrm{m}}{{\mathrm{L}}_{\mathrm{e}\mathrm{f}\mathrm{f}}}}\right)}^2+{\left({\displaystyle \frac{\mathrm{n}}{{\mathrm{W}}_{\mathrm{e}\mathrm{f}\mathrm{f}}}}\right)}^2}$$

(1)

where ε_eff is the effective relative permittivity of the patch, c is the speed of light in free space, and m and n are mode indices. The parameters L_eff and W_eff represent the effective length and width of the patch, respectively.

The resonant frequency f_mn at the fundamental TM10 mode can be determined using Equation (1). Additionally, the effective length L_eff is defined as follows:

$${\mathrm{f}}_{10}={\displaystyle \frac{\mathrm{c}}{2{\mathrm{L}}_{\mathrm{e}\mathrm{f}\mathrm{f}}\sqrt{{\epsilon }_{\mathrm{e}\mathrm{f}\mathrm{f}}}}}$$

(2)

$${\mathrm{L}}_{\mathrm{e}\mathrm{f}\mathrm{f}}=\mathrm{L}+2∆\mathrm{L}$$

(3)

The effective relative permittivity (ε_eff) accounts for the nonuniform electric field distribution between the patch and the ground plane and is determined using the equation provided by Schneider [22,23].

$${\epsilon }_{\mathrm{e}\mathrm{f}\mathrm{f}}={\displaystyle \frac{{\epsilon }_{\mathrm{r}}+1}{2}}+{\displaystyle \frac{{\epsilon }_{\mathrm{r}}-1}{2}}{\left[1+12{\displaystyle \frac{\mathrm{h}}{\mathrm{W}}}\right]}^{-{\displaystyle \frac{1}{2}}}$$

(4)

Due to the presence of fringing fields, there is difference between the electrical and physical dimensions of the antenna. The extension length ΔL caused by these fields can be estimated using the equation provided by Hammerstad [22,23].

$$∆\mathrm{L}=0.412 \mathrm{h}{\displaystyle \frac{\left({\epsilon }_{\mathrm{e}\mathrm{f}\mathrm{f}}+0.3\right)\left(\frac{\mathrm{W}}{\mathrm{h}}+0.264\right)}{\left({\epsilon }_{\mathrm{e}\mathrm{f}\mathrm{f}}-0.258\right)\left(\frac{\mathrm{W}}{\mathrm{h}}+0.8\right)}}$$

(5)

These extensions influence resonance and impedance matching, necessitating precise adjustments in the final design [24,25].

Adding slots or notches to a rectangular microstrip patch antenna changes the surface current paths and increases the antenna’s electrical length, creating additional resonant frequencies. This approach is widely used for dual-band or multiband operation [26].

Creating an upper gap at the radiation edge is particularly effective for achieving dual resonance. A compact microstrip patch antenna with inverted C-shaped and U-shaped slots functions as two connected resonators of different sizes. These slots guide meandering currents, generating multiple resonances. Dual-band operation can be achieved by strategically placing slots on the radiating patch, ground plane, or feed line, with the slot dimensions manually tuned to optimize performance [27].

Slots positioned at the radiation edge primarily influence higher-order modes, while center slots play a key role in generating a secondary frequency by extending the current path. This increases the effective resonant length and lowers the resonant frequency. Additionally, center slots suppress symmetric modes, affecting bandwidth and polarization [28].

The shape and placement of the slots significantly impact the antenna’s overall performance, influencing resonance, impedance matching, and polarization characteristics [23].

Notches placed near the feed arm of the rectangular patch influence surface currents and create new resonances. Their effect on the primary resonance is minimal due to their proximity to the feed arm. The depth of the bottom slot plays a critical role in forming a secondary resonance, while also affecting coupling coefficients and enabling impedance matching [29,30].

2.2. Simulation and Optimization

After determining the initial patch dimensions, the antenna design was modeled and simulated using CST Studio Suite, a widely recognized electromagnetic simulation tool that is extensively used in high-frequency antenna development. The primary objective was to achieve robust dual-band operation at 28 GHz and 38 GHz while optimizing critical performance parameters, such as S₁₁, gain, radiation efficiency, and directivity.

To enhance the accuracy of the simulation and account for real-world implementation factors, an SMP connector was integrated into the model using Antenna Magus software 2022 (https://www.3ds.com/products/simulia/antenna-magus, accessed date 16 February 2025). This addition enabled a more precise evaluation of the connector-induced effects, which are often overlooked but can significantly impact antenna performance at mmWave frequencies. The incorporation of the connector allowed for a more realistic assessment of impedance matching, signal integrity, and potential loss mechanisms.

During the optimization phase, various design parameters were systematically adjusted to improve the performance metrics. A detailed parametric analysis was conducted to examine the influence of key geometrical and material properties, including the following:

▪

Patch dimensions: The patch dimensions were modified iteratively to optimize the impedance matching and enhance the radiation efficiency.

▪

Substrate thickness and dielectric constant: Analyzed to balance gain enhancement with minimal losses, ensuring improved overall efficiency.

▪

Feed line configuration: Optimized for better impedance matching and reduced reflection coefficients.

▪

Ground plane modifications: These modifications were performed to fine-tune the radiation characteristics and improve the bandwidth.

An in-depth sensitivity analysis guided the iterative optimization process to evaluate the antenna’s resilience against minor manufacturing tolerances. This step is critical for ensuring minimal deviations due to fabrication inconsistencies. The final design exhibited significantly improved S₁₁ values, increased gain, and a more stable radiation pattern across the target frequency bands.

By combining CST-based full-wave simulations with targeted parametric optimizations and connector integration studies, this research presents a high-performance dual-band microstrip antenna capable of delivering superior results in next-generation wireless communication applications.

3. Results and Discussion

3.1. Antenna Design Overview

In this study, a rectangular microstrip patch antenna was selected due to its straightforward design and the ease with which its resonant frequency can be fine-tuned by adjusting the patch length (L). The patch dimensions were determined using standard microstrip antenna equations. Initially, a dual-band antenna for 28/38 GHz was designed, and its S₁₁ parameters were analyzed by modifying the geometry of the antenna.

The investigation involved generating multiple models and varying critical parameters, including the patch length, width, cavity dimensions, and the positioning of the patch arm. A seven-step sizing procedure was employed, where the patch length was incremented by 0.5 steps and the patch width was incremented by 1 mm, up to 20 mm. Based on the evaluation of the S₁₁ parameters, the optimal configurations were selected based on the gain, directivity, impedance matching, and voltage standing wave ratio (VSWR).

Additional customizations included repositioning the patch arm, modifying the gaps on the patch, and adjusting the dielectric and base materials. In some designs, the lower inserts on the patch arm were redesigned with alternative dimensions. Following these modifications, a parametric analysis was conducted to select four optimal designs based on S₁₁, gain, and VSWR.

The antenna design involved a comprehensive parametric analysis of the dielectric layer and patch dimensions. This analysis focused on key geometric factors, such as the gap dimensions (bw, bl), bottom gap dimensions (c, d), inner gap dimensions (e, f), and top gap dimensions (g, h). For all the designs, the patch width (yw) and length (yl) were fixed at 15 and 14 mm, respectively.

The insulator layer size was optimized to improve the antenna performance. After analyzing all the dimensions, the final antenna structure was developed, as illustrated in Figure 1.

The rectangular microstrip patch antenna underwent a detailed parametric analysis of the dielectric layer and patch dimensions to optimize performance. The key geometric aspects were systematically varied, such as gap widths, bottom gap lengths, and inner and top gap dimensions. These adjustments were essential for optimizing S₁₁ antenna operation and ensuring efficient antenna operation. The final antenna dimensions derived from this analysis are summarized in

Table 1. Antenna dimensions.

Parameter	Dimension (mm)	Parameter	Dimension (mm)	Parameter	Dimension (mm)
a	4.75	b	1.25	c	1
d	1	e	1	f	0.5
g	0.7	h	2.75	bw	1.25
bl	7	yw	15	yl	14
dw	30	dl	28	ykw	1
ykl	7

.

Figure 2 shows the simulation results of the S₁₁ parameters for the designed rectangular microstrip patch antenna across the specified frequency range. The simulation result presented in Figure 2 was generated using CST Studio Suite. Specifically, we utilized the Time Domain Solver, which employs the finite integration technique (FIT) to simulate the time evolution of electromagnetic fields. The FIT was selected for its high accuracy, efficiency, and proven numerical stability, as well as its flexibility in modeling complex geometries. Furthermore, the meshing process was conducted using CST’s automatic mesh generation algorithm, which produces a hexahedral mesh. The S₁₁ value, or return loss, is a critical parameter for assessing how well the antenna matches the transmission line, with lower values (below −10 dB) indicating better impedance matching and minimal power reflection.

The figure shows two prominent dips corresponding to the resonant frequencies at 26.6 GHz and 38.6 GHz, where the S₁₁ values drop well below −10 dB, reaching approximately −50 dB. These deep nulls indicate highly efficient energy transfer at these frequencies, which is consistent with the target dual-band operation of the antenna.

The graph also illustrates the antenna’s usable bandwidth, which is determined by the frequency ranges where S₁₁ remains below the −10 dB threshold. The 26.6- and 38.6-GHz bands exhibit sufficient bandwidth, confirming that the antenna operates effectively within these ranges with minimal reflection losses.

Figure 2 further confirms that the design adjustments, including patch sizing and geometric optimizations, resulted in strong impedance matching and high performance at both target frequencies. This confirms the antenna’s dual-band functionality. The extensive parametric analysis demonstrates the successful optimization of the antenna for both gain and impedance matching.

3.2. Connector Selection and Integration

The selection of appropriate connectors is paramount in ensuring optimal performance and minimizing variations in high-frequency antenna designs. As interconnection technologies advance, the focus is increasingly shifting toward extending the operational frequency ranges of antennas. A key factor in this development is understanding how the connector affects the antenna’s performance, particularly at high frequencies. Hence, choosing the most suitable connector type—such as the SubMiniature version A (SMA) or SubMiniature Push-on (SMP)—becomes essential for reducing signal loss and distortion.

SMA connectors are commonly used to link high-frequency circuits and antennas to receivers. Standard SMA connectors are typically rated for operation in the 18–24 GHz frequency range, depending on their construction and manufacturing quality. However, specialized versions, like 3.5 mm SMA connectors, are capable of handling frequencies as high as 26.5 GHz, with certain models even supporting frequencies as high as 30–35 GHz [20]. On the other hand, SMP connectors, which were introduced in the 1980s, have gained popularity due to their compact design and ease of connection, especially in high-density applications. SMP connectors support significantly higher frequencies, with some versions capable of operating up to 65 GHz [31].

For this study, a 2.92-mm SMP connector, rated for operation between 0 and 40 GHz, was integrated into the antenna design using Antenna Magus software to assess its effect on the overall performance. The choice of this connector was driven by its ability to meet the high-frequency demands of the dual-band design. The updated antenna model incorporating the SMP connector is shown in Figure 3.

This evaluation underscores the critical role of connector selection in high-frequency antenna design. Even slight mismatches in the connector performance can significantly impact key parameters, such as return loss and impedance matching. By selecting a connector with appropriate frequency-handling capabilities, the overall performance of the antenna system can be greatly enhanced.

Figure 4 presents a comparative analysis of the simulated S₁₁ parameters for the antenna design with and without the inclusion of an SMP connector, as assessed using the CST program. The blue curve illustrates the antenna’s performance without the connector, while the red curve depicts the S₁₁ response with the integrated SMP connector. In high-frequency structures, the manufacturing and connection precision of the antenna must be extremely high. This precision should also be minimized during the design process based on simulation results. To address this, the study explicitly demonstrates the impact of the connector on antenna design. Simulation results clearly indicate that the connector must be incorporated into the design for high-frequency applications. The introduction of the connector resulted in a noticeable shift in the central frequencies and a degradation in impedance matching, as evidenced by the higher S₁₁ values in the red curve. Considering these results, it has been conclusively shown that a structure without a connector would be unsuccessful in manufacturing, and therefore, production was not pursued.

The connector’s impact is particularly pronounced around the 28 and 38 GHz frequency bands, where the S₁₁ parameters show reduced performance. These results suggest that the presence of the connector causes an increase in signal reflection, leading to less efficient radiation and impedance matching. This comparison highlights the importance of meticulous connector selection and integration in high-frequency antenna design. Even small deviations in the connector performance can significantly alter the antenna’s operational characteristics, underscoring the need for precise engineering to optimize performance.

3.3. Optimization and Parametric Analysis

The optimization of the S₁₁ parameters for the connector-integrated antenna structure was performed using CST Studio Suite. To facilitate an efficient optimization process, the antenna model was restructured with a focus on key performance factors. The trust region framework (TRF) algorithm was selected due to its proven success in similar optimization studies reported in the literature. This optimization focused on fine-tuning the critical geometric parameters, including the top, side, and bottom gap dimensions, patch arm placement, and middle gap measurements. Constraints were applied to each of these parameters (ı, d, e, f, g, h, i, j, k, l, m7, n, o, p, r, s, t, u, v, y) to ensure precision.

After multiple iterations, the resulting S₁₁ values for the optimized designs are presented in Figure 5. Each curve represents a distinct optimized configuration (Opt-1, Opt-2, Opt-3, Opt-4), revealing how the parameter adjustments influence the return loss at the targeted 28 GHz and 38 GHz frequency bands. These results clearly demonstrate the efficacy of the parametric optimization process, leading to improved impedance matching and overall performance enhancement across the desired frequency bands.

After determining the optimized dimensions, further parametric analysis was conducted to refine the design and address potential challenges related to manufacturing tolerances. In this context, the spacing between the arms and the upper gap dimensions were identified as critical parameters due to their significant influence on S₁₁ and gain characteristics. The width (g) and height (h) of the arm spacing, along with the width (r) and height (s) of the upper gap, were systematically varied, starting at approximately 10% of the optimized step interval and progressively decreasing. These analyses were performed both independently and in combination to assess their impact on the overall design parameters. To account for manufacturing tolerances, the arm spacing parameters (g, h) were analyzed within the ranges of 0.4 mm–1.5 mm and 4 mm–8.5 mm, using step sizes between 0.15 mm and 0.5 mm. Similarly, the upper gap parameters (r, s) on the patch were examined with a step size of 0.1 mm across the intervals of 1.3 mm–2.1 mm and 1.9 mm–2.8 mm. The effects of these parametric variations on S₁₁ performance are illustrated in Figure 6. The simulation results indicate that the spacing between the arms has a more pronounced impact on the S₁₁ values, whereas the influence of the upper gap remains relatively limited.

Figure 7a,b show the S₁₁ results for the arm spacing (g and h) and upper gap (r and s) configurations. These investigations focused on fine-tuning the slot dimensions to minimize performance deviations due to fabrication inaccuracies, ensuring the antenna’s robustness and consistency throughout the specified frequency range.

3.4. Final Design Overview

Following the thorough parametric analysis, key antenna parameters, including the dielectric layer and the gaps along the sides, top, and arms (c, g, h, p, r, s), were fine-tuned to mitigate the effects of the manufacturing tolerances. After optimization, additional parametric analyses were conducted to determine the appropriate step spacing by considering antenna manufacturing sensitivity and the influence of each separation on the S₁₁ values. For the dielectric layer, which has minimal impact on S₁₁, a step interval of 0.5 mm was used, while intervals of 0.1 mm and 0.15 mm were applied for the upper gap analyses.

Building upon the optimized parameters, further investigations were conducted to analyze the impact of the slot dimensions on the dual-band operation of the proposed antenna. Since slot geometry plays a crucial role in resonance tuning and impedance matching, a detailed parametric study was performed to evaluate the influence of key slot dimensions. Figure 8 presents a comprehensive analysis of key slot dimensions affecting the dual-band operation of the proposed antenna. The parameters analyzed include the side slot width (p), side slot length (o), bottom slot width (l), and bottom slot length (m), detailed in Figure 9. The impact of these notches and slots on resonance frequencies, bandwidth, and return loss was systematically examined, leading to the identification of optimal slot configurations.

The side slot width (p) plays a significant role in influencing field distribution, impedance bandwidth, and resonance stability. The side slot length (o) is crucial for shaping the secondary resonance by modifying the effective inductance and capacitance. The bottom slot width (l) governs coupling strength, bandwidth, and radiation efficiency, while the bottom slot length (m) directly impacts the effective electrical path, affecting both impedance matching and resonance characteristics.

To systematically assess these effects, parametric analyses were conducted with a 10% step interval, while the notch structures were examined with a 100% dimensional variation to evaluate their full impact. The results depicted in Figure 8a,b indicate that variations in the side slot width (p) primarily affect the horizontal current distribution, leading to noticeable shifts in the lower resonance frequency. In contrast, the side slot height (o) has a more pronounced effect on impedance matching, as corroborated by both numerical and experimental findings. Furthermore, higher-order modes, which are generally more sensitive to vertical structural modifications, exhibit minimal influence on the secondary resonance frequency.

Figure 8c,d further analyze the effects of bottom slots on S₁₁. While these slots exert limited influence on current distribution, their proximity to the feeding point significantly impacts impedance matching. Consequently, even under substantial dimensional variations, their effect on resonance frequencies and bandwidth remains minimal, whereas their influence on return loss is found to be considerable, as validated through both numerical simulations and experimental measurements.

These findings underscore the critical role of slot geometry in achieving effective dual-band operation. The results emphasize the necessity of precise slot dimensioning to optimize impedance matching and radiation characteristics, thereby enhancing the overall antenna performance.

The resulting antenna model incorporating these optimized parameters is presented in Figure 9. The final design was selected based on a comprehensive S₁₁ and gain performance evaluation, ensuring the antenna meets the desired specifications. The detailed dimensions of the optimized antenna are summarized in

Table 2. Antenna dimensions of the final design.

Parameter	Dimension (mm)	Parameter	Dimension (mm)
a	16.50	l	1.40
b	13.50	m	0.55
e	13.75	o	0.70
f	17.68	p	4.45
g	0.83	r	2.05
h	5.41	s	2.35
i	1.55	u	0.76
j	0.60	y	1.63

, providing an in-depth breakdown of the critical measurements.

An S₁₁ graph was generated to assess the reflection performance across a broad range of frequencies after optimizing the antenna design. As shown in Figure 10, the graph clearly represents the antenna’s return loss over the operating frequencies, confirming that the design achieved significantly low S₁₁ values, particularly at the targeted frequencies. This validates the antenna’s efficiency and overall performance stability, thereby underscoring the success of the design optimization.

The simulation results reveal outstanding performance, with S₁₁ values as low as −53.04 dB at 28.08 GHz and −83.65 dB at 37.9 GHz. The antenna also maintained a return loss below −10 dB over two critical frequency bands: 27.63–28.75 GHz and 37.33–38.60 GHz, corresponding to bandwidths of 1.12 GHz and 1.27 GHz, respectively. These bandwidths are centered around the target frequencies of 28 GHz and 38 GHz, ensuring optimal operation across both bands. A comprehensive 3D model illustrating the optimized antenna configuration is presented in Figure 11, which presents the final design in greater detail. The presence of side lobes in the 3D radiation patterns is primarily attributed to the equal dimensions of the ground and insulator layers, along with the use of Rogers 5880 RT/duroid, a low dielectric material. This material was deliberately chosen to optimize S₁₁ and enhance gain performance at 28 GHz and 38 GHz. Although the design exhibits pronounced side lobes, it achieves a significantly higher gain than the literature average. Furthermore, side lobe levels can be effectively mitigated using established techniques, such as feed optimization, as documented in previous studies.

In terms of directivity, the antenna achieved an impressive value of 8.02 dBi at 28 GHz and 9.17 dBi at 38 GHz. The antenna also demonstrated exceptional efficiency, with total efficiencies of 96.02% at 28 GHz and 95.50% at 38 GHz, highlighting its superior operational performance. These results underscore the antenna’s ability to maintain high performance and efficiency across both frequency bands. A detailed depiction of the antenna’s overall realized gain and efficiency is provided in Figure 12.

Figure 13 illustrates the 2D radiation patterns of the designed antenna in the E-plane and H-plane at 28 GHz and 38 GHz. Figure 13a represents the E-plane radiation patterns at 28 GHz and 38 GHz, while Figure 13b shows the corresponding H-plane patterns, respectively.

3.5. Radiation Pattern Optimization

To optimize the antenna design, dual-band operation was achieved through a slot configuration, where the dimensions of each notch on the patch were systematically adjusted. Key antenna parameters, particularly S₁₁ and gain, were analyzed to determine the optimal design. In addition to this optimization process, extensive parametric analyses were conducted for each slot to further refine the radiation pattern. The optimization of the radiation pattern was carried out using two primary approaches: (a) ground plane modifications and (b) radiation pattern enhancement using EBG-like notches. These techniques enabled precise control over the beamforming characteristics, ultimately improving the antenna’s overall performance.

• Ground plane modifications

To assess the impact of ground plane modifications on the radiation pattern, parametric analyses were conducted by systematically varying the ground dimensions in increments of approximately 10% in all directions except, for the connector side. Instead of utilizing the CST Parametric Analysis tool, separate CST models were employed to minimize computational complexity and ensure accurate results. The simulation outcomes, including S₁₁, gain, and sidelobe level (SLL) for different ground plane configurations, are summarized in

Table 3. Parametric analysis of ground plane modifications.

Analysis Number	Ground Width (−x Direction)	Ground Width (+x Direction)	Ground Depth (+y Direction)	Ground Depth (−y Direction)	S11 (dB)		Resonant Frequencies (GHz)		Gain (dBi)		SLL (dB)
					28 GHz	38 GHz			28 GHz	38 GHz	28 GHz	38 GHz
1	15	16.5	13.5	13.5	−52.25	−37.41	28.09	37.92	7.47	8.24	−2.3	−7.1
2	13.5	16.5	13.5	13.5	−44.36	−43.07	28.12	37.96	7.62	7.76	−1.8	−6.8
3	12	16.5	13.5	13.5	−31.01	−32.97	28.14	37.98	7.83	8.02	−15.1	−1.7
4	16.5	15	13.5	13.5	−51.51	−36.46	28.09	37.93	7.47	8.24	−2.3	−7.1
5	16.5	13.5	13.5	13.5	−43.98	−43.45	28.12	37.96	7.62	7.76	−1.8	−6.8
6	16.5	12	13.5	13.5	−30.78	−32.30	28.14	37.99	7.83	8.02	−15.1	−1.7
7	16.5	16.5	12	13.5	−19.88	−33.84	28.32	38.56	6.23	8.48	−3.3	−2.9
8	16.5	16.5	10.5	13.5	−13.36	−26.48	28.43	38.73	6.38	8.93	−2.0	−2.5
9	15	15	13.5	13.5	−43.23	−43.45	28.13	37.96	7.62	7.76	−1.8	−6.8
10	15	15	12	13.5	−20.54	−22.22	28.39	38.63	6.22	7.69	−1.2	−1.2
11	15	15	10.5	13.5	−14.01	−18.94	28.47	38.66	6.18	8.28	−2.1	−1.0
12	13.5	13.5	13.5	13.5	−25.34	−27.53	28.13	37.95	7.72	8.30	−2.5	−2.5
13	13.5	13.5	12	13.5	−18.70	−24.85	28.44	38.49	6.37	7.62	−6.0	−1.4
14	13.5	13.5	10.5	13.5	−12.86	−23.81	28.63	38.57	6.24	7.83	−4.9	−2.5
Proposed Antenna	16.5	16.5	13.5	13.5	−53.00	−83.64	28.07	37.90	7.82	8.98	−2.2	−6.4

. The results demonstrate that variations in the ground plane dimensions significantly affect the antenna’s performance, particularly in terms of gain and the sidelobe level. The final design was selected based on an optimal trade-off between resonance stability, gain, and radiation pattern characteristics.

b.

Radiation pattern optimization using EBG-like notches

In addition to slot-based optimization and parametric analyses, electromagnetic bandgap (EBG)-like notches were symmetrically incorporated into the patch to enhance the radiation pattern, particularly by reducing the sidelobe levels. The notch dimensions, defined by parameters y and u, were systematically varied in increments of approximately 10% to evaluate their impact on the radiation characteristics. The primary objective of this optimization was to create a sufficient bandgap to suppress surface waves, thereby improving the overall radiation performance of the antenna. The simulation results for different notch configurations are summarized in

Table 4. Parametric analysis of radiation pattern optimization using EBG-like notches.

Analysis Number	Notch Parameter y	Notch Parameter u	S11 (dB)		Resonant Frequencies (GHz)		Gain (dBi)		SLL (dB)
			28 GHz	38 GHz			28 GHz	38 GHz	28 GHz	38 GHz
1	3.26	1.52	−19.45	−21.93	28.26	37.79	9.3	7.22	−15.7	−8.1
2	3.09	1.44	−19.54	−24.14	28.23	37.73	8.99	6.92	−15.6	−8.2
3	2.93	1.36	−20.77	−29.33	28.15	37.75	8.68	6.71	−15.5	−7.8
4	2.77	1.29	−21.93	−47.34	28.10	37.74	8.39	6.77	−15.5	−7.3
5	2.60	1.21	−23.60	−32.47	28.10	37.79	8.10	7.12	−15.6	−7.0
6	2.44	1.14	−25.18	−28.69	28.06	37.79	7.86	7.64	−1.1	−6.8
7	2.28	1.06	−27.95	−28.74	28.08	37.83	7.7	8.15	−1.5	−6.6
8	2.12	0.98	−30.36	−30.88	28.09	37.86	7.74	8.52	−1.8	−6.5
9	1.95	0.91	−35.12	−35.09	28.06	37.90	7.77	8.76	−2.0	−6.4
10	1.79	0.83	−41.72	−42.02	28.07	37.90	7.79	8.89	−2.1	−6.4
Proposed Antenna	1.63	0.76	−53.00	−83.64	28.07	37.90	7.82	8.98	−2.2	−6.4

.

To further optimize the notch design, additional parametric analyses were performed for smaller values of y and u. These analyses aimed to achieve an optimal balance between enhanced sidelobe suppression and minimal impact on return loss and gain. The final optimized dimensions were selected based on this trade-off, ensuring improved radiation performance without significant degradation in other key antenna parameters. The simulation results are summarized in

Table 5. Parametric analysis of radiation pattern optimization using smaller values of y and u for EBG-like notches.

Analysis Number	Notch Parameter y	Notch Parameter u	S11 (dB)		Resonant Frequencies (GHz)		Gain (dBi)		SLL (dB)
			28 GHz	38 GHz			28 GHz	38 GHz	28 GHz	38 GHz
1	1.43	0.66	−39.87	−42.82	28.08	37.89	7.83	9.02	−2.3	−6.5
2	1.43	0.56	−38.53	−42.02	28.08	37.90	7.82	9.03	−2.4	−6.4
3	1.43	0.46	−36.86	−39.70	28.08	37.90	7.81	9.03	−2.4	−6.6
4	1.43	0.36	−35.80	−39.63	28.08	37.88	7.82	9.03	−2.4	−6.6
5	1.23	0.66	−37.17	−39.63	28.08	37.88	7.85	9.03	−2.4	−6.6
6	1.23	0.56	−35.27	−40.00	29.09	37.90	7.84	9.03	−2.4	−6.6
7	1.23	0.46	−35.36	−38.04	28.07	37.88	7.84	9.03	−2.4	−6.6
8	1.23	0.36	−32.76	−39.34	28.06	37.88	7.85	9.03	−2.4	−6.6
9	1.03	0.66	−35.05	−37.77	28.09	37.90	7.86	9.03	−2.4	−6.6
10	1.03	0.56	−34.06	−39.36	28.07	37.88	7.86	9.04	−2.4	−6.6
11	1.03	0.46	−34.01	−38.75	28.09	37.88	7.86	9.04	−2.4	−6.6
12	1.03	0.36	−34.01	−39.04	28.09	37.88	7.86	9.03	−2.4	−6.6
Proposed Antenna	1.63	0.76	−53.00	−83.64	28.07	37.90	7.82	8.98	−2.2	−6.4

.

This study adopted a two-pronged approach to optimize the radiation pattern of the proposed antenna. Ground plane modifications were implemented to fine-tune the beamforming characteristics, while EBG-like notches were incorporated to suppress the sidelobes and enhance radiation efficiency. The results confirm that the proposed optimization methodology significantly improves antenna performance at both 28 GHz and 38 GHz, making it a promising candidate for 5G and applications beyond.

Figure 14 presents the null angles and gain values at 38 GHz. The purple dotted lines in the figure indicate the beamwidth considering the −3 dB threshold. The suboptimal radiation pattern at this frequency is due to the presence of nulls or sidelobes in certain directions. The antenna design incorporates a directed radiation structure to ensure stable and directional radiation. At 38 GHz, the radiation pattern achieves maximum gain at approximately 70°, with a narrow beamwidth of 27°, providing precise control. None of the nulls (located at 4°, 28°, 52°, 118°, 142°, 162°, and 178°) fall within the main lobe (approximately between 60° and 87°), so they do not impact system stability, which is governed by the main lobe. The CST simulation reports a sidelobe level of −6.4 dB at 38 GHz, even at the highest sidelobe, with the first sidelobe significantly lower than the main lobe level, as is commonly observed in the literature. This behavior, potentially caused by asymmetrical slot placement, is considered to be controlled. While the radiation pattern is not perfectly symmetrical, it remains highly functional.

The designed antenna’s radiation pattern at 38 GHz is specifically intended to fulfill the need for directional and compact radiation, critical for 5G mmWave applications. Since the antenna is directional and the nulls are situated far from the main radiation direction, they do not affect the effective coverage area or orientation. These nulls result naturally from resonance. For 5G and mmWave applications, directional, high-gain antennas are required due to free space losses, and omnidirectional radiation external antennas can be broadly used. This study emphasizes S₁₁ and gain values for optimal antenna design, and passive pattern optimization can be achieved with four different slot configurations, ground plane modifications, and EBG-like notch structures on the antenna patch. Parametric analysis of the antenna design enables this flexibility.

The antenna has a gain of 8.98 dBi at 38 GHz. The half-power beamwidth is 27°, and the maximum gain occurs at 71°. The −3 dB limits are approximately between 60° and 87°, with the nulls positioned outside the main radiation direction, as shown in Figure 14.

3.6. Fabrication and Testing

Building on the exceptional simulation results for S₁₁, gain, and efficiency, the single-element antenna demonstrated performance typically seen in array antennas. This impressive performance, surpassing many other designs in the literature, validated the decision to proceed with prototype fabrication. The fabrication was carried out using an LPKF laser machine, specially selected to handle the unique properties of the Rogers substrate material. The connector was meticulously integrated into the design once the antenna was fabricated, as shown in Figure 15.

A detailed comparison of the simulated S₁₁ values and those measured using a vector network analyzer (VNA) is presented in Figure 16, further confirming the consistency and high performance of the antenna across both analysis methods. The discrepancy observed in the S₁₁ value at the second resonance can be attributed to several factors related to the antenna fabrication and measurement process. Specifically, the sensitivity of the printed circuit board (PCB) during fabrication can lead to small variations in the material properties, which may affect the resonance characteristics. Additionally, tolerances in the connectors used and variations in the way the connectors are attached to the PCB can introduce slight differences in the impedance matching, further influencing the S₁₁ values.

3.7. Comparison with Existing Designs

Table 6. Comparison with similar published works.

Reference	Spectrum (GHz)	Bandwidth (GHz)	Antenna Elements	Antenna Size (mm2)	S11 (dB)	Gain (dBi)	Efficiency (%)	SLL (dB)
[12]	26–29.5	1.5	1 × 4 array	-	~−15	12.7	-	−15
	36.7–38.7	1.9			~−28	14.9		−12
[32]	25.59–29.63	4.04	Single	4.62 × 4.49	−19	3.5	94	-
	35.74–43.09	7.35			−26	5	95	-
[33]	27.10–28.80	1.7	2	4.8 × 2.9	−17	5.3–7.4	-	~−2
	35.20–38.90	3.7			−19	5–6.2	-	~−3
[34]	~27.50–28.75	1.25	Two-Port MIMO	7.5 × 8.8	−34.5	6.6	98.3	~−5
	~37.50–38.60	1.1			−27.3	5.86	98.5	~−3
[35]	~27.70–28.40	0.7	2 × 4 microstrip, dual-split-ring	130 × 53	~−35	4.6	55–60	~−7
	~36.50–39.50	3			~−24	10.5	88–90	~−8
[36]	27.00–28.40	1.4	Single/4	1.5 λ0 × 1.86 λ0	~−20 ~−20	6.73 12.69	94	-
	34.70–40.00	5.3			~−22 ~−20	5.51 11.1	94	-
[37]	~25–30	5	8	16 × 16	~−30 ~−50	15.6	-	~−2
	~36–42.5	6.2			~−20 ~−40	10	-	~−3
[38]	27.60–28.50	0.9	5 (Layer)	15 × 25	~−28	9	-	-
	36.90–38.90	2			~−22	5.9	-	-
[39]	25.4–30.84	5.44	1 × 4 array	6.5 × 23	~−23	11.31		−10
	38–40	2			~−22	11.93		−12
Proposed	27.63–28.75	1.12	Single	27 × 33	−53.04	7.82	96.02	−2.2
	37.33–38.60	1.27			−83.65	8.98	95.50	−6.4

presents a comparative analysis of various microstrip antennas designed for the 28/38 GHz frequency range. This table synthesizes data from recent studies, offering a comprehensive overview of the bandwidth, antenna configurations, physical dimensions, S₁₁ parameters, gain, and efficiency values.

The first two columns list each antenna’s operational frequency range (in GHz) and corresponding bandwidth (in GHz). The proposed design operates within the 27.63–28.75 GHz and 37.33–38.60 GHz ranges, achieving bandwidths of 1.12 GHz and 1.27 GHz, respectively. These values position the design competitively, though some existing works report larger bandwidths, up to 7.35 GHz, demonstrating the potential to balance bandwidth with gain.

The antenna configurations can vary from single-element designs to complex multi-element arrays. The proposed design, measuring 27 × 33 mm², is larger than certain single-element antennas, like those presented in [32], which measure as small as 4.62 × 4.49 mm². The physical dimensions are pivotal, as they influence the antenna’s performance, particularly the gain and efficiency.

The S₁₁ values, which indicate the return loss and the quality of impedance matching, are critical metrics for antenna performance. The proposed design exhibits remarkably low S₁₁ values of −53.04 dB at 28 GHz and −83.65 dB at 38 GHz, indicating excellent reflection minimization and optimal power transfer. In contrast, other studies reported S₁₁ values between −17 dB and −34.5 dB, highlighting the superior return loss achieved by the proposed design.

The gain (in dBi) is another crucial parameter that indicates the antenna’s ability to direct radio frequency energy. The proposed antenna achieved gains of 7.82 dBi at 28 GHz and 8.98 dBi at 38 GHz, outperforming other designs with lower gains. For instance, the designs reported by [33] exhibit gains ranging from 5.3 dBi to 7.4 dBi, with the proposed single-element antenna delivering comparable performance to multi-element arrays.

Efficiency, which measures how effectively an antenna converts input power into radiated energy, is another critical metric. The proposed design demonstrates total efficiencies of 96.02% at 28 GHz and 95.50% at 38 GHz—values that exceed the efficiencies of many comparable designs, where efficiencies range from 80% to 94%. The high efficiency underscores the practical viability of the proposed antenna in advanced communication applications.

SLLs in microstrip antennas are considered problematic because they can cause unwanted losses, interference, noise, and reduced directivity, all of which negatively affect antenna performance. The SLL of 5G microstrip antennas depends on various design factors, such as architecture, size, and intended application. An analysis of Table 3 shows that previously reported SLL values range from −2 dB to −15 dB. The proposed antenna achieves acceptable SLL performance at 38 GHz, matching well with values reported in the literature. However, the SLL of −2.2 dB at 28 GHz is higher than desired, indicating poor sidelobe suppression. This issue needs improvement since effective sidelobe suppression is essential for high directivity, reduced interference, and better signal quality in 5G systems.

In conclusion, the proposed antenna design offers an exceptional combination of low return loss, substantial gain, and high efficiency, outperforming several existing designs in the literature. This comparison not only validates the effectiveness of the employed design strategies but also underscores the potential for future advancements in antenna technology for the 28/38 GHz frequency bands.

4. Conclusions

The microstrip antenna designed and prototyped for the 28/38 GHz frequency bands, tailored for 5G and next-generation communication systems, represents a significant leap forward compared to existing designs in the literature. Initially, the design was simulated without connectors, enabling a focused assessment of the connector’s influence on the S₁₁ parameters post-integration. The subsequent optimization and parametric analyses played a pivotal role in refining the design, resulting in superior fabrication outcomes.

This single-element microstrip antenna delivers exceptional performance, with gains of 7.82 dBi at 28 GHz and 8.98 dBi at 38 GHz—surpassing typical antenna performance and approaching that of multi-element arrays.

The proposed antenna design achieves dual-band operation at 28 GHz and 38 GHz, with realized gain values of 7.82 dBi and 8.98 dBi, respectively. These gains are typical for single-element patch antennas. As observed in both the simulation and the VNA measurements, the remarkably low return loss further emphasizes the antenna’s excellence and distinct advantage over prior designs. Furthermore, the antenna achieves outstanding efficiency values of 96.02% at 28 GHz and 95.50% at 38 GHz while maintaining narrowband characteristics with bandwidths of 1.12 GHz and 1.27 GHz at their respective center frequencies.

The findings of this study reveal that while the proposed antenna achieves satisfactory SLL performance at 38 GHz, the −2.2 dB SLL at 28 GHz remains suboptimal, highlighting a need for improvement. Future research will focus on enhancing SLL performance at 28 GHz through advanced optimization techniques to achieve better sidelobe suppression and improve the antenna’s efficiency for high-performance 5G applications.

This study underscores the critical importance of optimization and parametric analysis in the development of high-performance microstrip antennas. This study highlights how a well-optimized single-element design can achieve both high gain and efficiency, demonstrating that smaller configurations can compete with more complex multi-element designs. The impact of the connector on antenna performance has been meticulously examined, providing valuable insights into how connector integration affects S₁₁ and overall performance. The results further illustrate how the antenna dimensions influence S₁₁, particularly through the parametric analysis, revealing the key design considerations for achieving optimal performance.