Miniaturized dual-band MIMO antenna with high gain and isolation for mm-wave applications

Abstract

This work presents a compact dual-band four-element multiple-input multiple-output (MIMO) antenna designed for millimeter-wave (mm-wave) applications, with particular emphasis on 5G wireless communication systems. The proposed antenna occupies an ultra-compact volume of 15 × 15 × 0.8 mm³ (approximately 0.55λ × 0.55λ × 0.03λ at 28 GHz), making it highly suitable for integration into space-constrained wireless devices. The antenna operates efficiently over two mm-wave bands, 29–31 GHz and 36.5–38.5 GHz, covering key 5G NR-FR2 frequency allocations. By incorporating strategically placed slots and a compact radiating structure, the antenna achieves a peak realized gain of 8.1 dBi and 8.64 dBi in the lower and upper operating bands, respectively, while maintaining strong port isolation exceeding 25 dB. Both simulated and measured results demonstrate good agreement, validating the effectiveness of the proposed design. A comprehensive MIMO performance evaluation confirms excellent diversity characteristics, with a very low envelope correlation coefficient (ECC) of less than 0.001, a high diversity gain (DG) of approximately 9.99 dB, a total active reflection coefficient (TARC) maintained within − 5 to 5 dB, and a low channel capacity loss (CCL) of 0.21 bit/s/Hz across the operating bands. Owing to its compact size, high gain, strong isolation, and robust MIMO performance, the proposed antenna is a promising candidate for next-generation mm-wave 5G and future wireless communication systems.

Introduction

Millimeter-wave (mm-wave) communication has become a key enabler for fifth-generation (5G) and future wireless systems due to its capability to support ultra-high data rates, enhanced spectral efficiency, and low-latency transmission. In particular, the frequency bands around 28 GHz and 38 GHz have been widely allocated for 5G NR-FR2 deployments, enabling multi-gigabit wireless links for enhanced mobile broadband applications. However, mm-wave signals are highly susceptible to free-space path loss, blockage, and rapid channel variations, which severely limit the reliability and coverage of single-antenna systems. To overcome these limitations, multiple-input multiple-output (MIMO) technology is extensively employed, as it improves channel capacity, link robustness, and spectral efficiency through spatial diversity and multiplexing. By utilizing multiple antenna elements within a compact footprint, MIMO systems mitigate fading effects and improve overall communication reliability, making them indispensable for modern mm-wave wireless devices.

Among the various antenna platforms, microstrip antennas are particularly attractive for mm-wave MIMO implementations due to their low-profile planar structure, light weight, ease of fabrication, and compatibility with standard printed circuit board (PCB) technology. Recent studies have demonstrated that microstrip-based antennas can achieve reliable mm-wave performance when appropriate modeling, material selection, and dimensional optimization are applied^31–33. Advances in microstrip antenna analysis and modeling have further enabled accurate prediction of resonant behavior and impedance characteristics at mm-wave frequencies, making them suitable for compact and high-frequency MIMO designs^34,36. In addition, recent compact and planar antenna configurations have shown that microstrip antennas can effectively support multi-element MIMO arrangements with improved gain and radiation stability, while remaining cost-effective and scalable for mass production^35,37.

To address the stringent performance requirements of mm-wave MIMO systems, numerous antenna design strategies have been reported in the literature. Compact dual-band mm-wave antennas with improved isolation for 5G applications have been demonstrated in early works^1,2. Techniques such as defected ground structures (DGS) and electromagnetic bandgap (EBG) configurations have been widely employed to suppress surface waves and reduce mutual coupling between closely spaced antenna elements^3,4. Other approaches, including polarization diversity and novel feeding techniques, have contributed to enhanced isolation and impedance matching^5,6. Compact antenna arrays and metamaterial-inspired designs have also been explored to improve gain and bandwidth performance, although these approaches often increase structural complexity or antenna size^7–10. Subsequent studies have refined these concepts through optimized compact geometries and improved isolation mechanisms^13–15. More advanced solutions employing metasurfaces, parasitic loading, and coplanar decoupling structures have achieved excellent isolation and radiation performance, but typically rely on complex or bulky configurations^21–24. Recent quad-port, flexible, fractal, and integrated sub-6 GHz/mm-wave MIMO antennas have further extended multi-band and multi-port capabilities; however, these designs generally occupy larger footprints or require sophisticated isolation mechanisms that limit their suitability for compact user equipment^38–41.

In contrast, the proposed work introduces a highly compact dual-band four-element microstrip MIMO antenna with an overall size of 15 × 15 × 0.8 mm³, achieving strong inter-element isolation, competitive gain, and excellent MIMO performance without employing complex decoupling structures. This improved size-to-performance trade-off clearly differentiates the proposed antenna from existing designs and highlights its suitability for next-generation compact mm-wave 5G and future wireless communication devices.

Major contributions:

1. Novel Compact Dual-Band MIMO Antenna Design: The paper presents a new and innovative compact dual-band four-element MIMO antenna design that achieves high gain, high isolation, and a compact size of 15 mm x 15 mm x 0.8 mm.

2. High Gain and Isolation: The antenna design achieves a peak gain of up to 8 dBi over the operating spectrum while maintaining excellent isolation of over 25 dB, making it suitable for mm-wave applications.

3. Compact Size and Reliability: The antenna’s compact size, high performance, and reliability make it an optimum choice for mm-wave applications, including 5G wireless communication systems.

4. Simplified Design and Fabrication: The design’s simplicity and use of conventional materials make it more feasible for fabrication and integration in small form factor devices, reducing the complexity and cost of production.

5. Innovative Structural Changes: The incorporation of strategically placed slots in the radiating element is a novel approach that enables the antenna to achieve high gain and isolation while maintaining a compact size.

These contributions highlight the significance and impact of the paper’s research, demonstrating a novel and effective solution for mm-wave antenna design that addresses the needs of 5G wireless communication systems.

Antenna design and analysis

Figure 1 illustrates a proposed antenna system structure consisting of four antennas on a shared substrate. In Fig. 1(a), the overall arrangement of the four antennas is shown within a square region with dimensions (W) (width) and (L) (length). The green background represents the substrate, a dielectric material that provides physical support and influences the electrical performance of the antennas. The brown regions indicate the metallic parts of each antenna, which serve as the radiating elements responsible for transmitting and receiving electromagnetic waves. The antennas are labeled as Antenna 1, Antenna 2, Antenna 3, and Antenna 4, each positioned strategically across the substrate to reduce interference and optimize signal isolation, which is essential in multi-antenna systems for better diversity and multiplexing gains. In Fig. 1(b), a detailed view of Antenna 1 is provided, revealing its intricate structure and various dimensions. This close-up shows the antenna’s geometry with labeled dimensions, including widths ((W1), (W2), (W3), etc.) and lengths ((L1), (L2), (L3), etc.). These measurements are essential for defining the resonant frequency, impedance matching, and radiation efficiency of the antenna. The design incorporates nested metallic sections with various trace widths and lengths, suggesting that the antenna is engineered to support multi-band or wideband functionality, enabling it to operate efficiently across different frequency ranges. Each section of Antenna 1’s layout plays a specific role in determining the current distribution and electromagnetic field patterns. For instance, the longer segments may contribute to resonance at lower frequencies, while shorter segments can be tuned for higher frequencies, effectively making it a multi-resonant structure. The narrow gaps and varied trace widths indicate that precise impedance matching techniques have been applied to maximize power transfer and minimize signal reflection. This configuration is well-suited for applications requiring multiple antennas on a single substrate, such as in MIMO systems, where multiple signals are transmitted and received simultaneously for improved data rates and reliability. The substrate and the positioning of each antenna indicate a deliberate attempt to enhance performance by managing inter-antenna coupling and ensuring sufficient isolation between the elements.Table 1 provides various design parametres of the proposed MIMO antenna.

Fig. 1.

(a) Overall structure of proposed antenna, (b) deatiled dimension of antenna 1.

Table 1.

Various design parameters of proposed MIMO antenna.

Parameters	Values (mm)	Parameters	Values (mm)
L	15	W2	0.28
L1	7.4	W3	1.3
L2	2.6	W4	0.46
L3	3.3	W5	0.28
L4	0.6	W6	0.6533
L5	0.3	W7	0.14
L6	1.5	W8	0.188
W	15	W9	3
W1	1.12	W10	0.2

The proposed MIMO antenna is implemented on a Rogers RT/duroid 5880 substrate with a relative permittivity of ε_r = 2.2, a loss tangent of tanδ = 0.002, and a thickness of 0.8 mm. This substrate was selected due to its low dielectric constant and very low loss tangent, which are essential for minimizing dielectric losses and maintaining high radiation efficiency at millimeter-wave frequencies. The low permittivity also helps to reduce surface-wave propagation, thereby contributing to improved impedance matching and enhanced isolation between closely spaced MIMO elements. The chosen substrate thickness represents a trade-off between mechanical stability, bandwidth enhancement, and suppression of higher-order modes, making it suitable for compact mm-wave antenna designs.

The radiating patch and feed lines are realized using copper (σ ≈ 5.8 × 10⁷ S/m), which offers high electrical conductivity and low ohmic losses at high frequencies. Copper is widely employed in microstrip antenna fabrication due to its excellent current-carrying capability, ease of etching, and compatibility with standard printed circuit board (PCB) manufacturing processes. The ground plane is also formed using copper and extends beneath the substrate to provide a stable reference plane and consistent current return path. The use of a full copper ground plane enhances radiation stability, reduces unwanted back radiation, and supports reliable impedance characteristics across the operating bands.

Overall, the combination of a low-loss dielectric substrate and high-conductivity metallic layers ensures efficient radiation, stable dual-band performance, and strong MIMO characteristics while maintaining a compact and fabrication-friendly structure. The selected materials also facilitate practical realization and seamless integration into mm-wave wireless devices.

Although the spacing between the antenna elements in Fig. 1 may appear relatively large, it was carefully optimized considering the electrical wavelength at millimeter-wave frequencies and the stringent isolation requirements of compact MIMO systems. At the operating bands around 30 GHz and 38 GHz, the free-space wavelength is relatively small, and the selected edge-to-edge spacing corresponds to approximately 0.3λ–0.4λ, which is sufficient to significantly suppress near-field coupling while still maintaining a compact overall footprint. This spacing minimizes surface-wave interaction and reduces mutual coupling without requiring additional decoupling structures such as defected ground, electromagnetic bandgap, or metasurface layers. The effectiveness of the chosen spacing is validated by the measured isolation exceeding 25 dB and the very low envelope correlation coefficient (ECC ≈ 0.001), confirming that the spacing strategy plays a key role in achieving the desired MIMO performance.

The design of the proposed microstrip-based MIMO antenna follows well-established analytical formulations commonly used for initial dimension estimation and resonance prediction. The effective permittivity (​) of the substrate, given in (1), accounts for the fringing fields extending into the air region and is widely employed in modern microstrip antenna modelling to accurately estimate wave propagation characteristics at millimetre-wave frequencies^31,36. The guided wavelength and quarter-wavelength relations expressed in (2) and (3) are fundamental to determining resonant behaviour and feed-line dimensions in planar antenna structures and remain valid for compact mm-wave designs when combined with full-wave optimization^32,34,42.

The expression for the patch width in (4) is used to control the input impedance and radiation efficiency of the antenna and provides a first-order estimate for achieving efficient radiation at the target frequency^31,33. Fringing field effects, which cause the effective electrical length of the patch to exceed its physical length, are incorporated through the length extension formulation in (5), while the effective resonant length given in (6) ensures excitation of the dominant mode^32,36. These formulations are commonly adopted in recent microstrip antenna studies and remain effective for compact and high-frequency implementations when followed by numerical optimization.

For antennas incorporating slots and modified radiating geometries, the resonant frequency is governed by the effective current path length, as expressed in (7), which is widely used in compact and multi-resonant antenna designs^34,35. The characteristic impedance relation in (9) enables proper feed-line matching to a standard 50-Ω source and is essential for minimizing reflection losses in mm-wave circuits^33,37. Finally, the inter-element spacing criterion given in (10) is selected to balance compactness and mutual coupling suppression in MIMO configurations, ensuring acceptable isolation and stable radiation performance^35,36,42.

Although these closed-form expressions provide useful initial estimates, the final antenna dimensions were obtained through full-wave electromagnetic simulation and parametric optimization to account for coupling effects, higher-order modes, and fabrication tolerances inherent to compact mm-wave MIMO structures.

Effective permittivity (Hammerstad):

1

where is the dielectric constant of the substrate, is its thickness, and is the width of the patch. This equation indicates that as the substrate becomes thinner or the patch becomes wider, approaches ​, whereas for very wide substrates it tends toward unity.

The propagation constant inside the substrate is determined by the guided wavelength:

2

where is the velocity of light and is the operating frequency. For many design purposes, a quarter-wave resonant length is required, particularly for feed lines or resonant stubs, which is given by:

3

The width of the patch is a critical parameter influencing the input impedance and radiation efficiency. It is approximated as:

4

where ​ is the fundamental resonant frequency. A wider patch tends to increase bandwidth and reduce input impedance.

Due to fringing fields at the edges of the patch, the physical length appears electrically longer than its actual dimension. The extension in length is expressed as:

5

The effective length of the patch required to support the fundamental TM10 mode is,

6

where ​ is the free-space wavelength. This ensures the resonant condition is satisfied.

For antennas incorporating slots, the resonant frequency is determined by the effective slot length .

7

where ​ is the effective permittivity seen by the slot.

In the case of a square ring resonator, the dominant mode resonates when the guided wavelength equals the perimeter ,

8

he characteristic impedance of a microstrip line is crucial for proper impedance matching. For >2, Hammerstad’s approximation gives:

9

This allows the feed line to be designed for a standard 50 Ω match.

In multi-element or MIMO antenna systems, inter-element spacing governs mutual coupling and array performance. A practical range is,

10

This ensures minimal grating lobes while maintaining acceptable isolation between elements.

Figure 2 illustrates the reflection and transmission coefficients of the proposed MIMO antenna, emphasizing its dual-band performance and inter-element isolation. In Fig. 2(a), the reflection coefficients (S11, S22, S33, S44) show two prominent dips around 30 GHz and 37 GHz, aligning with the green-shaded operating bands. These dips indicate effective impedance matching at these frequencies, suggesting the antenna is optimized for dual-band operation. Figure 2(b) presents the transmission coefficients (S-parameters (S(ij))between the antenna pairs. The low transmission values across the operating bands, largely below − 20 dB, demonstrate excellent isolation among antennas, which is crucial for high-efficiency MIMO performance by minimizing inter-antenna coupling. This design achieves the necessary isolation and frequency selectivity to support advanced communication systems within the targeted frequency bands.

Fig. 2.

(a) Reflection coefficient, (b) transmission coefficient of the proposed MIMO antenna.

The parametric analysis shown in Fig. 3 examines the impact of varying three design parameters—top strip length (L6), rectangular strip length (L7), and slot length (L8)—on the reflection coefficient (S11) spanning frequencies from 26 to 40 GHz. In Fig. 3 (a), changes in L6 primarily affect the impedance matching in the 33–35 GHz range, with the proposed length of 1.501 mm yielding optimal results within the target bands (highlighted in green) at 28 GHz and 35 GHz. In Fig. 3 (b), variations in L7 similarly influence the reflection coefficient, particularly around the 28 GHz and 36 GHz points, where the proposed length of 0.2813 mm demonstrates better matching within the desired operating bands. Figure 3 (c) shows the effect of adjusting L8, with the proposed value of 0.3697 mm delivering improved impedance matching over the two target frequency bands. These results collectively indicate that the selected dimensions for L6, L7, and L8 contribute to stable dual-band performance, ensuring that the antenna maintains low reflection coefficients (below − 10 dB) across the designated frequency ranges, which is crucial for effective millimeter-wave 5G operation.

Fig. 3.

Parametric analysis when altering (a) top strip length (L6), (b) rectangular strip length (L7), (c) slot length (L8).

Figure 4 illustrates the design evolution of the proposed antenna through four stages and shows how each stage affects the reflection coefficient (S11). In Step 1, the basic square patch and single feed structure result in poor impedance matching across the desired frequency bands. Step 2 introduces dual feeding strips, which slightly improves impedance matching, especially around the 28 GHz range, but the performance at higher frequencies remains suboptimal. Step 3 adds an inner square loop, enhancing the resonance around 34 GHz and improving overall impedance matching; however, the dual-band performance is not yet optimized. In the final proposed design, a nested inner structure is added, resulting in significant improvement in S11 values within the target frequency bands (28 GHz and 35 GHz), as indicated by the deep dips in the reflection coefficient below − 10 dB. This final configuration achieves stable dual-band operation by effectively suppressing reflection across both bands, thus demonstrating the effectiveness of each structural modification in achieving the desired dual-band performance for millimeter-wave 5G applications.

Fig. 4.

(a) Various evolution stages of proposed antenna and their (b) reflection coefficient.

Dual band operating principle

The proposed antenna achieves dual-band operation through controlled modulation of the surface current paths introduced by the embedded slot geometry within the radiating element. At the lower operating band (around 30 GHz), the dominant resonance is associated with a longer effective current path formed along the outer edges of the radiator and the connected strip structure (Figure. 5(a)), resulting in increased electrical length and lower resonant frequency. Conversely, the upper operating band (around 38 GHz) is generated by a shorter current path created by the inner slot and compact strip sections (Figure. 5(b)), which reduce the effective electrical length and shift the resonance to a higher frequency. This behavior is validated by the surface current distributions and parametric analysis, where variations in slot and strip dimensions independently tune the lower and upper resonances. Consequently, the antenna realizes stable dual-band performance within a single-layer, compact microstrip structure without requiring additional resonators or complex decoupling elements.

As seen in the surface current distribution plots (Fig. 5), the current at each frequency (30 GHz and 37 GHz) is primarily concentrated around the active port with minimal spillover or excitation of nearby elements, suggesting low coupling between ports. This behaviour indicates that the antenna geometry and the placement of resonating elements reduce unwanted coupling, maintaining a high level of isolation.

Fig. 5.

Surface current distribution at (a) 30 GHz, (b) 37 GHz.

Measured results and discussions

The simulated MIMO antenna was fabricated and is shown in Fig. 6(a) and (b). Each antenna element is connected to a 50Ω SMA connector, which is then measured using a Keysight Vector Network Analyzer (VNA). During the measurement process, the SMA connector of the antenna being tested is connected to the VNA. In contrast, the remaining connectors are terminated with a 50Ω load to prevent any unwanted reflections. Figure 7 presents the measured S-parameters of the proposed antenna, providing valuable insights into its performance. The reflection coefficient plot (Fig. 7(a)) shows that the antenna exhibits good impedance matching across the frequency range of interest, with all S-parameters (S11, S22, S33, S44) remaining below − 10 dB, indicating a low reflection coefficient and efficient radiation. The transmission coefficient plot (Fig. 7(b)) reveals the coupling between different antenna elements, with S-parameters (S12, S23, S34, S13, S14, S24, S32, S42) displaying varying levels of transmission. The low transmission coefficients (< −20 dB) indicate minimal coupling between elements, suggesting good isolation and reduced interference. Overall, the measured S-parameters demonstrate the antenna’s potential for efficient and isolated operation, making it suitable for MIMO applications.

Fig. 6.

Fabricated prototype of the proposed antenna (a) front view, (b) back view.

Fig. 7.

Measured S-parameters of the proposed antenna, (a) reflection coefficient, (b) transmission coefficient.

Figure 8 (a) – (d) illustrates the far-field radiation patterns of two antennas at 30 GHz and 37 GHz, with separate analyses for E-phi and E-theta components. Antenna 1, depicted in Fig. 8 (a) & (b), and antenna 2, shown in 8 (c) & (d), reveal distinct lobe formations at each frequency, indicating directional radiation behaviour that supports focused signal transmission. Across both frequencies, the antennas exhibit stable and predictable radiation patterns with main lobes and side lobes characteristic of high-gain designs, ideal for minimizing interference and enhancing signal strength in targeted directions. Minor discrepancies between simulated and measured results suggest minimal impact from fabrication tolerances, while the consistency of the patterns across both antennas emphasizes their suitability for high-frequency applications where precise directivity and reliable beam shaping are critical for efficient data transmission and coverage.

Fig. 8.

Far-field pattern of antenna 1 at (a) 30 GHz, (b) 37 GHz, antenna 2 at (c) 30 GHz, (d) 37 GHz.

Figure 9 compares the simulated and measured gain of Ant 1 and Ant 2, highlighting the impact of fabrication tolerances and substrate effects on antenna performance. The discrepancies between simulated and measured gain curves, particularly in the upper-frequency range, suggest that the model may not fully capture the complex interactions between the antenna elements and the surrounding environment. The measured gain of Ant 1 exhibits a notable ripple effect, indicative of potential substrate modes or edge diffraction effects that are not fully accounted for in the simulation. In contrast, Ant 2’s measured gain shows a smoother frequency response, implying a more effective suppression of unwanted modes.

Fig. 9.

Simulated and measured gain of the proposed antenna.

MIMO performance

The MIMO performance of the proposed antenna is evaluated using the envelope correlation coefficient (ECC), diversity gain (DG), total active reflection coefficient (TARC), and channel capacity loss (CCL). ECC quantifies the correlation between antenna elements, where values below 0.5 are acceptable and values below 0.1 indicate excellent diversity performance. DG represents the improvement in signal reliability due to diversity and is inversely related to ECC, with values close to 10 dB considered desirable. TARC characterizes the combined reflection behaviour of all ports under simultaneous excitation and provides a realistic measure of impedance matching in MIMO systems; values within approximately − 5 to 0 dB are generally acceptable. CCL measures the reduction in theoretical channel capacity due to correlation and mutual coupling, with values below 0.4-bit/s/Hz indicating efficient MIMO operation. These definitions and performance criteria are consistent with standard MIMO antenna analysis reported in recent studies^43,44. Figure 10(a–d) illustrates the MIMO performance of the proposed antenna, including its diversity gain, channel capacity Loss (CCL), envelope correlation coefficient (ECC), and total active reflection coefficient (TARC). Figure 10 (a) reveals that the antenna combinations exhibit low ECC values, indicating good diversity performance, with values below 0.5 across the frequency range. The diversity gain plot in Fig. 10 (b) demonstrates that the antenna combinations achieve a gain of around 10 dB, signifying effective diversity performance. Figure 10 (c) shows that the antenna orientations have a minimal impact on the received channel, with TARC values ranging from − 5 to 5 dB. Finally, Fig. 10 (d) indicates that the channel capacity loss is relatively low, with values below 1 bps/Hz, suggesting that the proposed antenna design can efficiently utilize the available channel capacity. Overall, the figure suggests that the proposed antenna design offers excellent MIMO performance, making it suitable for applications requiring high data rates and reliable communication.

Fig. 10.

MIMO performance of the proposed antenna, simulated and measured (a) ECC, (b) diversity gain, (c) TARC, (d) CCL.

Compared with the reported works summarized in Table 2, the proposed antenna demonstrates a more favourable balance between compactness and MIMO performance, particularly for mm-Wave dual-band operation. In terms of physical size, the proposed design occupies only 15 × 15 × 0.8 mm³, which is significantly smaller than most existing antennas, including recent quad-port and multi-band designs (Refs. 21–30, 38–41), many of which exceed 20–30 mm in lateral dimensions or employ volumetric substrates. Despite this substantial miniaturization, the proposed antenna achieves competitive dual-band gains of 8.1 dB and 8.64 dB at 30 GHz and 38 GHz, respectively, which are comparable to or higher than several larger counterparts operating at similar frequencies. Moreover, the antenna maintains robust port isolation of 28 dB, satisfying practical mm-Wave MIMO requirements while avoiding complex decoupling networks or bulky isolation structures used in some prior works. From a diversity standpoint, the proposed design exhibits excellent MIMO metrics, with a very low ECC of 0.001 and a high diversity gain of 9.99 dB, outperforming or matching most reported designs, including recent multi-band and flexible MIMO antennas. In addition, the channel capacity loss (0.21-bit/s/Hz) remains well within acceptable limits, confirming efficient multi-stream transmission. Overall, the proposed work offers a superior size-to-performance trade-off, making it particularly well suited for compact mm-Wave terminals and next-generation 5G/6G devices, where space constraints and high MIMO efficiency are simultaneously critical.

Table 2.

Performance comparison of the proposed antenna with literature.

Ref.	Size (mm³)	Frequency of operation (GHz)	Gain (dBi)	Isolation (dB)	DG	ECC	CCL (bit/s/Hz)
21	20 × 40 × 1.6	28	7.8	29.34	9.96	0.05	Not given
22	26 × 14 × 0.762	28 and 38	6.2 and 5.9	28	9.99	0.06	Not given
23	80 × 80 × 12	6	8	15.5	9.98	0.004	Not given
24	28 × 16 × 6.3	40	10	33	9.98	0.1	Not given
25	26 × 14.5 × 0.508	28 and 38	5.2 and 5.5	39 @ 28 GHz, 38 @ 38 GHz	9.99	0.0001	0.05
26	30 × 15 × 0.25	28	5.42	35.8	9.99	0.005	0.1
27	20 × 40 × 1.6	28	14.1	35	–	0.01	Not given
28	30 × 35 × 0.76	28	8.3	10	9.96	0.05	0.4
29	28 × 28 × 0.79	28 and 38	11.65 and 13.65	47.85	9.99	0.001	0.1
30	70 × 40 × 0.787	4.2 and 4.9	Not given	25	9.96	0.005	Not given
38	~ 24 × 24 × 0.8	28 and 38	9.2 and 9.8	> 30	9.98	0.002	Not given
39	~ 30 × 30 × 0.5	26/28/38	6.5–8.1	> 25	9.97	0.01	Not given
40	~ 22 × 22 × 0.8	28	7.4	> 20	9.95	0.02	Not given
41	~ 45 × 45 × 1.6	Sub-6 & 28	6.8 (mmWave)	> 22	9.96	0.01	0.3
Proposed work	15 × 15 × 0.8	30 and 38	8.1 and 8.64	28	9.99	0.001	0.21

Conclusion

This work has presented a compact dual-band four-element microstrip MIMO antenna designed for millimeter-wave applications, particularly targeting 5G NR-FR2 frequency bands. The proposed antenna achieves dual-band operation over 29–31 GHz and 36.5–38.5 GHz within a highly compact footprint of 15 × 15 × 0.8 mm³, making it well suited for integration into space-constrained wireless devices. Despite its small size, the antenna delivers competitive realized gains of 8.1 dBi and 8.64 dBi in the lower and upper operating bands, respectively, while maintaining strong inter-element isolation exceeding 25 dB. Comprehensive MIMO performance evaluation confirms excellent diversity characteristics, including a very low envelope correlation coefficient (ECC ≈ 0.001), high diversity gain (≈ 9.99 dB), low total active reflection coefficient (TARC within − 5 to 5 dB), and a low channel capacity loss (0.21 bit/s/Hz). The close agreement between simulated and measured results validates the robustness and practical feasibility of the proposed design.

Future work may extend this antenna concept toward larger MIMO configurations, such as massive MIMO arrays, or adapt the geometry for reconfigurable and beam-steerable mm-wave systems. Additionally, the proposed design can be explored for conformal and wearable platforms, as well as for emerging 6G frequency bands, by leveraging its compact and planar architecture. Owing to its favorable size-to-performance trade-off, the antenna is a strong candidate for applications in 5G user equipment, small-cell base stations, vehicular communications, and next-generation wireless devices.

Author contributions

All authors contributed meaningfully to the development of this research work. R. Gayathri led the conceptualization of the antenna design, carried out the simulation studies, and coordinated the overall manuscript preparation. K. Kavitha provided technical guidance on MIMO principles, supported the refinement of the design methodology, and contributed to result interpretation and manuscript editing. Rajesh Kumar D assisted in fabricating the prototype, conducting measurements, and validating the simulated performance with experimental results. Sundaravadivel P supervised the research process, ensured methodological rigor, and contributed to the final revision and quality improvement of the manuscript. All authors reviewed and approved the final version of the paper and agree to be accountable for its contents.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.