2 × 2 MIMO dual-wideband ground radiation antenna with a T-shaped isolator for Wi-Fi 6/6E/7 applications

Abstract

In this paper, a miniaturized 2 × 2 MIMO dual-wideband ground radiation antenna targeting Wi-Fi 6/6E/7 standards using 2.4 GHz, 5 GHz, and 6 GHz frequency bands with sufficient antenna performance was designed. The proposed antenna system contains four identical 4 mm × 6 mm antennas of the internal loop type and two identical 6 mm × 6 mm isolators containing lumped LC elements. This paper considers maximizing the impedance bandwidth using an advanced impedance control technique and isolation enhancement through T-shaped isolators. To achieve dual-wideband resonance, an impedance control method using LC parameter value control is employed by utilizing the coupling between the ground plane, the voltage source, the feed loop, and the radiation loop in each radiation band. The operating principle of the T-shaped isolator was proposed using the scattering matrix and impedance matching. Simulated results showed that the proposed antenna can cover 96 MHz at the 2.4 GHz band and 2,232 MHz at both the 5 GHz and 6 GHz bands with an impedance bandwidth of -10 dB and isolation performance of less than − 20 dB across all frequency bands. Average measured efficiencies of 52%, 74%, and 80% were obtained for the corresponding operating bands.

Introduction

The continued growth of the Internet of Things (IoT) and the demand for high-bandwidth applications have led to the need for more advanced wireless networking technologies. Wi-Fi 6 (802.11ax), Wi-Fi 6E, and Wi-Fi 7 (802.11be) are designed to meet the growing demand for higher data rates and increased network capacity. These standards are expected to provide superior wireless communication with significantly improved performance in dense network environments compared to previous Wi-Fi standards^1,2. One of the key areas of improvement in Wi-Fi 6/6E/7 is the use of the 2.4 GHz, 5 GHz, and 6 GHz frequency bands. For this reason, several studies involving various approaches for antenna applications using all three bands have been proposed to meet increasing demands. These antennas include multiple-input multiple-output (MIMO) patch^3–5, open cavity⁶, low-profile laptop^7,8, square-shaped modified monopole⁹, loop and monopole laser direct structuring¹⁰, and two-port microstrip¹¹ antennas targeting Wi-Fi 6/6E and 7. However, a miniaturized antenna suitable for hand-sized portable IoT devices covering all three bands has yet to be proposed.

In the mobile communication industry, the size limitations of the mobile antenna have always been a challenging issue. To address this concern, a ground radiation antenna has been proposed to satisfy the requirement of a small-sized internal antenna inserted in a printed circuit board that maintains high antenna performance¹². The underlying concept of this approach is focused on controlling magnetic coupling between the miniaturized radiation loop and the whole electronic device by analyzing the current distribution of the ground characteristic mode. This approach enables the overall device to function as a huge resonator, resulting in a dramatic improvement in radiation performance. Additionally, the antenna impedance matching process proceeds without requiring an extra fabrication step during the actual design process, which is a significant advantage. For these reasons, research on the ground radiation method has evolved to provide viable mobile antenna design technologies^12–20.

In addition, a dual-band ground radiation antenna covering the 2.4 GHz and 5 GHz wireless local-area network (WLAN) bands has been developed²⁰. The previous study presented a dual-band radiation method that controls two resonant loop structures sharing a single-loop-type feed structure. The method focuses on the operating frequency and input impedance tuning on each band to maximize ground excitation. Specifically, a sufficient impedance matching process that controls the coupling between the feed loop and the radiation loop was proposed. However, the proposed method cannot satisfy the bandwidth requirements of Wi-Fi 6/6E/7 applications. To satisfy the requirements of the given standard, impedance bandwidth improvement through an additional impedance control technique is needed.

In this paper, a dual-wideband operation technique using a ground radiation antenna with additional impedance control targeting the Wi-Fi 6/6E/7 standards with sufficient antenna performance is proposed. The proposed antenna is composed of a feed loop and two radiation loops for the corresponding radiating band. The control of the coupling between the magnetic field induced by the feed loop and the field source representing the voltage source using lumped LC components was analyzed in this study to produce extra impedance bandwidth enhancement. Consequently, wideband operation, which can be controlled by the given lumped LC element with a fixed antenna structure, is available through the supplementary method. In other words, wideband resonance can be obtained by controlling the input impedance circle on a Smith chart within the − 10 dB magnitude circle. The proposed approach can reduce the antenna clearance size to 4 mm × 6 mm while performing at 96 MHz at the low band and 2,232 MHz at the high band impedance bandwidth; this is satisfactory radiation performance for portable devices using Wi-Fi 6/6E/7 applications.

In order to continue the evolution of Wi-Fi technology towards Wi-Fi 6/6E/7, which represents a significant leap in wireless communication capabilities, the full potential of these advancements requires continued research into MIMO antenna technology as well as the high radiation performance of a single antenna²¹. In particular, 2 × 2 MIMO applications where two antennas are used at both the transmitter and receiver is widely adopted in Wi-Fi applications because it provides a significant improvement in data throughput and link reliability while maintaining a manageable cost and complexity^22–24. In MIMO antenna applications, the isolation between the antennas is important for radiation performance because insufficient isolation can lead to increased signal correlation between antennas, reducing the effective capacity of the MIMO system. By improving isolation, MIMO antennas can maintain low correlation, enabling higher data rates and more reliable communication. This is particularly important for advanced Wi-Fi standards that aim to deliver extremely high throughput (EHT) and support complex features like multi-link operation (MLO). For these reasons, isolation enhancement through dollar-shaped strip²⁵, rectangular stub²⁶, Y-shaped stub^27,28, inverted L-shaped short strip²⁹ and boat-shaped structure³⁰ methods have been proposed to achieve more than − 20 dB isolation for dual-band antennas. In this paper, a 2 × 2 MIMO dual-wideband ground radiation antenna in a hand-sized ground plane with advanced decoupling performance provided by a T-shaped loop type isolator (6 mm × 6 mm) containing series lumped capacitors is proposed. The resonance frequency of the isolating structure can be varied by tuning the series capacitor, which is a significant advantage in the design procedure in actual applications.

In this paper, a 2 × 2 MIMO dual-wideband ground radiation antenna with a T-shaped isolator is proposed to satisfy the radiation performance requirements of the given applications. The noticeable contributions presented in this paper are listed as follows:

• Antenna input impedance bandwidth enhancement through a source coupling control loop to achieve the impedance bandwidth requirements of the given standard.

• Isolating performance enhancement through a T-shaped decoupling structure to achieve − 20 dB isolation between antennas.

• The above issues discussed and analyzed via simulation and measurement.

The proposed antenna design offers several key highlights that make it suitable for modern Wi-Fi 6/6E/7 applications. First, the antenna achieves dual-wideband operation, covering 2.4 GHz, 5 GHz, and 6 GHz frequency bands, through optimization of magnetic coupling between the antenna structure and the ground plane. This ensures consistent performance across multiple wireless communication standards. Second, the T-shaped isolator effectively enhances isolation performance by suppressing mutual coupling between antenna elements, maintaining isolation below − 20 dB across all frequency bands. Additionally, the design is highly compact and efficient, with the antenna occupying only 4 mm × 6 mm and the isolator requiring just 6 mm × 6 mm. This miniaturized configuration makes the design ideal for space-constrained devices while maintaining the performance levels necessary for commercial applications. These features collectively demonstrate the practicality and innovation of the proposed antenna design.

Results

Antenna design

Figure 1 illustrates the geometry of the proposed 2 × 2 MIMO dual-wideband antenna and a T-shaped isolator in an overall view, as well as the specific antenna and isolator configuration. Figure 2 illustrates its current distribution. The optimized size definition and lumped inductor and capacitor values are given in Table 1, and the thickness of the copper path is 0.5 mm. As shown in the figure, four antennas are positioned 10 mm away from the short edge and two isolators are positioned at the center of the long side. Isolator #1 decouples antennas #1 and #2, and isolator #2 isolates antennas #3 and #4. The ground radiation antenna is miniaturized with a 4 mm × 6 mm clearance size, and it is sufficient to excite a 100 mm × 50 mm ground plane. The side containing ​ in the antenna structure is positioned facing outward of the ground plane. The size of the ground plane is set for small modern electronic devices, and it is printed on a 1-mm-thick FR4 substrate. Antenna simulations were carried out using the ANSYS high-frequency structure simulator (HFSS) tool using the driven modal solution type.

Fig. 1.

Geometry of the proposed antenna: (a) overall configuration, (b) antenna configuration, and (c) isolator configuration.

Fig. 2.

Simulated surface current distributions: (a) with isolator at 2.4 GHz, (b) without isolator at 2.4 GHz, (c) with isolator at 5 and 6 GHz, and (d) without isolator at 5 and 6 GHz.

Table 1.

Design parameters of proposed antenna.

Parameter	Value	Parameter	Value	Parameter	Value

The dual-wideband operation of the proposed MIMO antenna is achieved by optimizing the magnetic coupling between the antenna elements through precise control of LC lumped components and their interaction with the ground plane. By controlling the inductance and capacitance values, the antenna’s resonant frequencies and input impedance are tuned to cover the desired Wi-Fi 6/6E/7 frequency bands, ensuring efficient wideband performance. Furthermore, the T-shaped isolator, designed with strategically placed LC lumped elements, enhances isolation by suppressing mutual coupling between the antenna elements. This isolator operates as a decoupling structure by suppressing mutual coupling between antenna elements, thereby significantly reducing interference and enhancing MIMO performance. Together, these design strategies ensure optimal dual-wideband operation and enhanced isolation performance.

Figure 2 illustrates the current distributions of the overall antenna configuration. It is operated by four port excitations in each antenna. The operating principle of the ground radiation antenna is based on creating strong magnetic coupling between the voltage source, feed loop, radiating loop, and ground plane, effectively turning the ground plane into a large dipole antenna^12–20. According to the reaction concept of the ground characteristic mode analysis, the magnetic coupling between the magnetic field produced by the ground plane, , and the induced magnetic current density, , yields the ground radiation antenna operation. This can be expressed as follows^31,32:

1

It is important to analyze the dominant mode excitation, which is the long side of the ground plane radiating as a dipole-type antenna. Therefore, the ground radiation antenna should be positioned where the magnetic coupling between the ground plane and the antenna is strongest. The length of the long side, , is set to be of 6 GHz and of 2.4 GHz. It is noted that the current distribution of every points is the null position, which leads to the weakest coupling, while the point leads to the greatest coupling. The center of the antenna clearance is positioned from the edge side to achieve strong coupling, where it is for high-band operation and for low-band operation.

Figure 3 depicts the current distributions of the proposed antenna in each radiation band. Similar to the ground radiation antenna method, the current induced by the feed loop, including and , excites the radiation loop of each band^12–20. In low-band operation, a parallel loop containing , , and is excited to cover the 2.4 GHz band, where the loop size and lumped LC values control the resonance frequency. The input impedance can be matched by controlling the magnetic coupling between the feed and radiation loops. In high-band operation, the series loop containing and is that of radiation, which can be operated in both the 5 GHz and 6 GHz bands. The size of the low-band radiation loop can be reduced using large lumped inductor and capacitor values, and even a smaller resonance loop can be formed than with the high-band radiation loop. The operation mechanism can be identified through a lumped LC parameter study.

Fig. 3.

Simulated surface current distributions of antenna at (a) 2.4 GHz and (b) 5 and 6 GHz.

Figure 4 shows the current distributions of the proposed T-shaped isolator. The operation mechanism of the isolator is based on suppressing the coupling between two antennas by creating a loop-type resonance structure of 2.4 GHz and 6 GHz containing a series capacitors and . Each capacitor is designed to control the resonance frequency of the isolating loop, where the resonance frequency of the loop resonator can be expressed as follows³³:

Fig. 4.

Simulated surface current distributions of isolator at (a) 2.4 GHz and (b) 5 and 6 GHz.

2

where and are the resonance frequencies of each isolator, and are the series inductance values of the proposed loop, and and are the series capacitance values that are inserted to match the resonance frequency. Also, it is noted that the isolator does not operate as a radiating antenna because the coupling between the ground plane is negligible, where it is positioned in the null zone of the current distribution of the ground plane.

Figure 5 depicts the simulated S parameters. The port numbers correspond sequentially to the antenna numbers. For the reflection coefficient of antennas, the simulation results demonstrate 96 MHz at the 2.4 GHz band and 2,232 MHz at the 5/6 GHz band of the − 10 dB impedance bandwidth. Unlike previous dual-band antennas, an ultra-wideband frequency bandwidth is obtained, which exceeds the specifications of the given standard. The overall wideband operating principles are introduced in the method section. Regarding antenna coupling coefficients, additional isolation improvement is not required because the magnitude of is already under − 20 dB. For , about − 11 dB at low band and − 13 dB at high band is represented as coupling coefficient without the isolator. However, with the proposed T-shaped isolator, the coupling coefficient is enhanced to achieve a magnitude under − 20 dB for all frequency bandwidths. The decoupling procedure is organized in the isolator design method section. is not varied by the existence of the isolator.

Fig. 5.

Simulated S parameters.

Measurement results and radiation performance comparison

In this section, antenna performance verification through a bare board design is discussed. An antenna prototype was produced to measure the reflection coefficient, coupling coefficient, radiation patterns, and radiation efficiency in the Wi-Fi 6/6E/7 operating bands. As an advantage of the ground radiation design method, it is easy to conduct impedance matching and optimization using LC value tuning without extra fabrication. The reflection coefficient, gain, efficiency, and radiation patterns are measured using a vector network analyzer in a 3D anechoic chamber.

Figure 6 shows photographs of the measurement setup and the proposed antenna design. The measurement is conducted in a chamber with a receiving horn antenna to analyze the far-field radiation performance. The handmade prototype antennas are designed with and without the isolator to compare the radiation performance. The antenna prototype is made in a 1-mm-thick FR4 substrate with 0.1-mm copper tape. The antenna design parameters use similar values to the data represented in Table 1.

Fig. 6.

Measurement setup and antenna prototype photos: (a) 3D anechoic chamber with receiving horn antenna, (b) transmitting proposed antenna in chamber, (c) overall antenna view without isolator, (d) overall antenna view with isolator, (e) specific single antenna, and (f) specific isolator.

Figure 7 shows a reflection coefficient comparison of the simulated and measured data. The overall tendency of the measured antenna performance follows the simulated data. An approximate minimum of 2,439 MHz and maximum of 2,678 MHz − 10 dB bandwidth is achieved, which is a wider bandwidth than the simulated data.

Fig. 7.

Measured and simulated reflection coefficient data comparison.

Figure 8 illustrates the simulated and measured peak realized gain and total efficiency of a single antenna for each radiation band. The gain (G) of the antenna is calculated by incorporating the directivity (D) and the total efficiency () of the antenna, as given by the equation:

Fig. 8.

Simulated and measured peak antenna gain and antenna efficiency.

3

In the simulation results, the average peak realized gain is observed to be 2.32 dBi, 2.62 dBi, and 2.98 dBi for the 2.4 GHz, 5 GHz, and 6 GHz bands, respectively. The highest simulated peak gain of 3.39 dBi is observed at 6.85 GHz. These values demonstrate consistent performance across the desired frequency bands and align with the expected values for commercial-grade antennas. The total efficiency () is derived as the ratio of the radiated power (​) to the input power (​​), which can be expressed as follow:

4

In the simulation, the average total efficiency is achieved as 52%, 74%, and 78% for the 2.4 GHz, 5 GHz, and 6 GHz bands, respectively. This indicates effective power transfer and low losses in the antenna design, even at the bare-board design stage. Additionally, the highest efficiency of 80% is observed at 6.6 GHz, highlighting the antenna’s optimized performance in the higher frequency band. These results confirm that the antenna achieves high gain and efficiency, making it suitable for applications targeting Wi-Fi 6/6E/7 standards.

Figure 9 presents the radiation patterns of a single antenna for each operating frequency band. Similar to a dipole-type antenna, omnidirectional radiation patterns are obtained for every band where the operating principle of the ground radiation antenna is exciting the huge conducting body as dipole antenna. It depicts the radiation characteristics in the xy-plane and yz-plane for each frequency band. The graphs are decomposed into the -polarization component, -polarization component, and the total electric field. the θ-polarization component dominates in the xy-plane, exhibiting omnidirectional radiation characteristics. The -polarization component is relatively weaker compared to the component, indicating a strong vertical polarization radiation. In the yz-plane, the radiation pattern exhibits left-right symmetry with minimal radiation at 0° and 180°, which can be attributed to the strong coupling effect with the ground plane. Overall gains of 2.32 dBi, 2.61 dBi, and 2.62 dBi for 2.45 GHz, 5.5 GHz, and 6.5 GHz, respectively, are obtained for the XY plane. Gains of 2.39 dBi, 2.59 dBi, and 2.62 dBi for 2.45 GHz, 5.5 GHz, and 6.5 GHz are observed for the XZ plane, respectively. These results demonstrate adequate performance for mobile antenna applications.

Fig. 9.

Measured radiation patterns on the XY and XZ planes at 2.45 GHz, 5.5 GHz, and 6.5 GHz.

Figure 10 shows the measured coupling coefficients containing and in with and without the isolator and , , and . The straight lines and dotted lines shows a similar result because these are in a symmetrical structure. Overall tendency follows the simulated data and every coupling coefficient are represented in under − 20 dB isolation performances as well.

Fig. 10.

Measured S parameters.

Table 2 presents comparison data of the proposed antenna and other relevant antennas for Wi-Fi 6/6E or 7 applications. Although the proposed antenna has a relatively large overall size of 100 × 50 mm compared to other antennas, the actual functional areas consist of four 4 × 6 mm regions for radiation and two 6 × 6 mm regions for isolation. The remaining space can be utilized for the electronic components necessary for IoT operation, making this design highly advantageous for hand-sized IoT devices, such as mobile phones or even smaller electronic devices. Moreover, its convenient impedance matching process can proceed using only lumped LC value tuning without extra fabrication. Compared to the previous dual-band ground radiation antenna, high band impedance bandwidth improvements of 720 MHz to 2,232 MHz have been achieved, where the proposed model effectively satisfies the performance requirements of Wi-Fi 6/6E and 7 standards with further reductions in antenna clearance size²⁰.

Table 2.

Wi-Fi 6/6E/7 antennas performance comparison with references.

References	Antenna Type	Operating Band	Antenna Size ()	Low Band Frequency Bandwidth (GHz)	High Band Frequency Bandwidth (GHz)	Peak Gain (dBi)	Peak Radiation Efficiency (%)
3	MIMO Patch Antenna	Dual				5.8	86
4	Dielectric Patch Antenna	Single		NA		8.47	NA
5	MIMO Patch Antenna	Single		NA		12.05	72.7
6	Open Cavity Antenna	Single		NA		7.8	95
7	Monopole	Dual				4.6	78
8	Monopole	Dual				3	68.6
10	EMC Hybrid Loop/Monopole LDS Antenna	Dual				5.7	77
20	Ground Radiation Antenna	Dual				2.12	71
This Work	Ground Radiation Antenna	Dual				3.42	80

Table 3 compares the isolation values of different MIMO antennas that can be applied to Wi-Fi 6/6E/7 applications. Other references, such as hybrid metal walls, dollar-shaped strips, rectangular stubs, and L-shaped short strips, have been used to enhance isolation performance^3,6,25,26,29. These methods require additional external components, necessitating complex procedures for improving isolation performance in actual applications. However, the proposed method is an efficient solution suitable for mobile IoT devices, as it enhances isolation under − 20 dB for given frequency bandwidths by using only a 6 mm × 6 mm clearance area on the printed circuit board.

Table 3.

Wi-Fi 6/6E/7 isolation enhancement performance comparison with references.

References	Antenna type	Operating band	Antenna Size ()	Low band frequency bandwidth (GHz)	High band frequency bandwidth (GHz)	Decoupling structure	Isolation (dB)	ECC
3	MIMO Patch Antenna	Dual				Hybrid Metal Walls	> 20	< 0.01
6	Open Cavity Antenna	Single		NA		NA	> 21.4	< 0.05
25	Monopole	Dual				Dollar-shaped Strip	> 15	< 0.02
26	Trapezoidal Monopole	Dual				Rectangular Stub	> 15	< 0.01
29	Monopole	Dual				L-shaped Short Strip	> 19.3	< 0.03
This Work	Ground Radiation Antenna	Dual				T-shaped Branch in Embedded Slot	> 20	< 0.01

Method

Antenna impedance control method

For the Wi-Fi 6/6E/7 standard, approximately 4% of the 2.45 GHz bandwidth and more than 30% of the 6 GHz band must be obtained to satisfy the bandwidth requirements. Thus, wideband optimization in the high-band process must be achieved while maintaining low-band resonance. In other words, the high-band resonance inside the − 10 dB circle on the Smith chart shown in Fig. 11 is the main objective to achieve wideband operation for the given application.

Fig. 11.

Simulated input impedance of the proposed antenna on a Smith chart.

Previous research has focused on designing an impedance control technique using the coupling between the feed loop and the radiation loop in dual-band operation²⁰. However, the proposed antenna cannot cover the frequency bandwidth requirement of the given standard. To solve this problem, both the input impedance control of the feed structure and the radiating loop structure must be analyzed to realize input impedance bandwidth enhancement.

The operation mechanism of the high-band radiation relies on the induced current of the feed loop exciting the radiating loop resonance, which is similar to the characteristics of the voltage source of circuit theory. Hence, the coupling between the source and the feed loop determines the input admittance of the feed loop resonance, which can be written as follows³¹:

5

where is the input voltage, is the field source representing the voltage source, and is the magnetic field induced by the voltage source. As observed in Eq. (5), the input admittance of the feed loop can be controlled by the coupling of the source magnetic current and the induced magnetic field in the constant input current value. Hence, if the coupling increases, the admittance will also increase. The same result will be produced for the opposite case. On the Smith chart, the magnitude of the input admittance of the feed loop determines size of the circle locus, which starts from the right end and moves in the clockwise direction.

Since the admittance of the coupling is too strong in the given feed loop, an extra decoupling loop must be formed. With the additional decoupling branch containing and , as shown in Fig. 12, the feed loop rotation current is hindered; this decreases the coupling between the magnetic current and magnetic field. Accordingly, the feed loop resonance can form into a --10 dB magnitude circle to achieve wideband performance.

Fig. 12.

Equivalent circuit in high-band operation.

As a results, the wideband impedance bandwidth can be obtained through controlling the coupling between the feed loop and the radiation loop using the additional control loop. However, it is difficult to balance the resonance frequency with a sufficient impedance bandwidth for both radiation bands because the LC components have both coupling and frequency variations. Thus, a parameter study of each LC component must be conducted to establish the dual wideband ground radiation antenna for Wi-Fi 6/6E/7 application.

Lumped LC parameter study

In this section, variations of the significant LC values ( of the proposed antenna are analyzed by focusing on the input impedance and resonance frequency of the respective feed and radiation loops. Each element affects the antenna performance, and the predominant factor is known. Understanding the tendency of each LC element toward the objective is necessary to design dual wideband antennas in various types of ground planes. Since the two radiation bands are affected differently by the LC parameters, the impedance and resonance frequency change tendencies of the operating bands should be analyzed separately through a Smith chart. The black lines in the figures are the reference simulated data, where the parameters are listed in Table 1. The parameters not mentioned in the legend of the figure are shown in Table 1.

Figure 13 depicts the simulated input impedance variation when is varied. When the value increases, the high-band resonance circle moves clockwise but the coupling between the high-band radiation loop and the feed loop is not affected much as it is contained in both loops. However, the impedance at the resonance point decreases because the magnetic coupling from the feed loop and voltage increases. In other words, the loop current flows more to the feed loop instead of the source coupling loop. In low-band operation, the coupling between the feed loop and the low-band radiation loop is greater since the resonance frequency of the feed loop gets closer to the low-band resonance frequency.

Fig. 13.

Simulated input impedance on a Smith chart with different values.

Figure 14 depicts the simulated input impedance variation when varies. Unlike the variation, the magnetic coupling between the source and the feed loop is not affected much. Also, the coupling between the high-band radiation loop and the feed loop are affected. In low-band operation, the overall tendency follows the variation, including the feed-loop resonance point and input impedance. However, the amount of difference varies so that the magnitude of the LC value can control the input impedance while maintaining the resonance frequency.

Fig. 14.

Simulated input impedance on a Smith chart with different values.

Figure 15 illustrates the simulated input impedance variation when varies. The resonance frequency of the high band is directly affected, whereas the feed loop is scarcely affected. When the magnitude of increases, the resonance frequency of the high band decreases and moves counterclockwise, which is opposite that of . Moreover, the amount of the impedance circle movement is different, which means that the ratios of , , and can change the location of the circle while maintaining the size and resonance frequency of the circle. Variation of does not affect the low-band operation much, which proves that the low-band radiation mostly operates in the loop containing , , and .

Fig. 15.

Simulated input impedance on a Smith chart with different values.

Figure 16 depicts the simulated input impedance variation when varies. When increases in high-band operation, the current flowing to the source coupling loop increases, which causes a decrease in the coupling between the feed loop and the field source. This results in the feed loop impedance controlling the wideband impedance bandwidth. When increases in low-band operation, the resonance frequency decreases with little difference in the coupling between the feed loop.

Fig. 16.

Simulated input impedance on a Smith chart with different values.

Figure 17 depicts the simulated input impedance variation when varies. The parameter mechanism is similar to in low-band operation, but it is not affected much in the high-band operating mode. Since the high-band radiation loop contains and in series, the difference in is not very significant.

Fig. 17.

Simulated input impedances on a Smith chart with different values.

Figure 18 depicts the simulated reflection coefficient with different and values for the fixed resonance frequency of the low-band radiation loop at 2.4 GHz. The proportion of and can change the high-band resonance frequency locus while maintaining the low-band resonance frequency, and vice versa.

Fig. 18.

Simulated reflection coefficient with different and values.

In this study, the LC parameter variations (​, ​, ​, ​, ​) of the proposed antenna were systematically analyzed to understand their effects on the input impedance and resonance frequency for dual-wideband operation. Each parameter significantly influences the performance of the antenna, with distinct tendencies observed in low- and high-band operations. The study highlights that the interaction of these parameters allows precise control over the resonance frequencies and impedance bandwidth while preserving the overall antenna structure.

Key insights include the observation that variations in ​ and ​ affect both the low- and high-band operations by shifting the resonance frequency and impedance characteristics, with magnetic coupling playing a pivotal role. ​, in contrast, directly influences the high-band resonance frequency with minimal impact on the low band, demonstrating its isolated effect on high-band performance. ​ was shown to control the wideband impedance bandwidth in the high band by altering the coupling between the feed loop and source coupling loop, while ​ primarily affected low-band operation with limited influence on the high band. Furthermore, the study demonstrated that the proportion of ​ and ​ can adjust the high-band resonance frequency without affecting the low-band resonance, showcasing the flexibility of LC parameter tuning.

Through this comprehensive analysis, it is evident that the proposed antenna’s impedance characteristics can be effectively tailored to adapt to complex ground structures in modern electronic devices. The ability to fine-tune wideband impedance bandwidth via LC parameter variation ensures adaptability to different operating environments and supports robust dual-wideband performance. This approach paves the way for efficient antenna design for compact and high-performance communication devices.

Isolator design

In this section, the design process and performance analysis of a T-shaped isolator, capable of maintaining the isolation performance of all 2 2 MIMO dual wideband antennas below − 20 dB, are discussed. Since the coupling coefficients between antenna #1 and antennas #3 and #4 already exhibit performance below − 20 dB, isolation enhancement is not addressed for these pairs. The focus is on improving isolation between antennas #1 and #2 across both 2.4 GHz and 5/6 GHz bands, where an identical method is adapted between antennas #3 and #4. Suppose antennas #1 and #2 are ports 1 and 2 and isolator #1 is port 5. Here, the voltage relationship between proposed antennas #1 and #2 and isolator #1 can be expressed in a three-port network system, as follows³⁴:

6

where and represent the incident and reflected voltage from port , represents the reflection coefficients of port , and represents the coupling coefficient between ports and . According to the decoupling theorem of a microwave network system, we can express the equation of influence on the correlation of antennas by an isolator, as follows³⁴:

7

where and are the coupling coefficients from antenna #1 and #2 with and without an isolator, and is the reflection coefficient of the isolator at the termination condition, which can be expressed as follows³⁴:

8

where is the load impedance and the characteristic impedance. According to Eqs. (7) and (8), we can show that coupling coefficient between antennas #1 and #2 can be controlled by the impedance matching of the proposed isolator. Thus, the variation tendency according to the design parameters must be analyzed through simulation results. For all simulation results, parameters not specified in the legend follow the values provided in Table I.

Figure 19 illustrates the variation in across each band as a variation of ​. It can be observed that, as the value of increases, the resonant frequency of the isolator in the 2.4 GHz band decreases. In contrast, a decrease in results in an increase in the resonant frequency. It is crucial to precisely design the resonant frequency to be set at 2.4 GHz. Changes in the 5/6 GHz band with varying ​ are found to be minimal.

Fig. 19.

Simulated coupling coefficient with different values.

Figure 20 shows the variation in across each band as a variation of ​. Variation in the resonant frequency with varying capacitance values in the 5/6 GHz band is evident, while changes in the 2.4 GHz band are negligible. The analysis of the variation tendencies of and confirms that each isolator loop operates independently, necessitating accurate design based on suitable capacitance values. Additionally, it is essential to understand these variation trends, as the resonant frequency of the loop can be affected by the designed location and surrounding electronic components.

Fig. 20.

Simulated coupling coefficient with different values.

Figure 21 shows the variation in as a function of the isolator height ​. Resonant frequency changes due to height variations were adjusted using the values of ​ and . It can be observed that, as the default height (or the depth of the slot) increases, the isolation performance improves. This is indicated by the wider frequency bandwidth where S₂₁ is enhanced. When the depth is 4 mm, the isolator fails to meet the − 20 dB threshold in the higher frequency band, whereas the best isolation is achieved at a depth of 8 mm.

Fig. 21.

Simulated coupling coefficient with different values.

Figure 22 illustrates the variation in as a function of the isolator width ​. In this case as well, the resonant frequency was matched by adjusting and ​. Similar to the tendency observed with height, a wider width results in better isolation performance. When the width is reduced to 4 mm, the isolator fails to achieve the − 20 dB isolation. When designing mobile antennas for portable IoT devices, space constraints are a significant consideration. Reducing the size of the isolator provides a substantial economic advantage, and appropriate design techniques must be thoroughly researched.

Fig. 22.

Simulated coupling coefficient with different values.

The envelope correlation coefficient (ECC) is an important parameter of MIMO antenna systems. In order to calculate the ECC by the simulated S parameters, we can use the following Eqs^10,35:

9

10

where and are the complex conjugate values of and and and represents the 3-dimensional far-field radiation pattern. Figure 23 shows the simulated and measured of the proposed MIMO antenna system. It is noted that the simulated ECC value is less than 0.01 for every radiating band of the modified system in both simulated and measured results. This result indicates excellent performance characteristics for a mobile MIMO antenna embedded in the ground plane.

Fig. 23.

Simulated and measured envelope correlation coefficient.

Conclusion

A dual-wideband 2 × 2 MIMO ground radiation antenna for Wi-Fi 6/6E and 7 applications with enhanced isolation performance was proposed. We designed a miniaturized loop-type antenna for a dual-wideband antenna covering the 2.4 GHz, 5 GHz, and 6 GHz bands. It thus satisfies the required frequency bandwidth of the given standard. To further fulfill the requirements, wideband optimization through an impedance control method was analyzed, and results of 96 MHz at the 2.4 GHz band and 2,232 MHz at the 5/6 GHz band of input impedance bandwidth are obtained with isolation performances of under − 20 dB between every antenna. The suggested method can be applied to various types of wireless communication devices using impedance tuning with lumped LC elements without extra fabrication.

Author contributions

J.K. wrote the paper and performed the simulations. J.K., H.S., and S.W. addressed the antenna impedance enhancement methods. J.K., H.L. and R.L. fabricated the proposed antenna and conducted measurements. H.K. organized the manuscript and revised the entire work. All authors contributed to the writing of the manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.