A Flower-Shaped Miniaturized UWB-MIMO Antenna with High Isolation

An ultra-wideband (UWB) multiple-input, multiple-output (MIMO) antenna with a reason-ably compact size of 30 × 18 × 1.6 mm 3 is presented in this paper. The proposed antenna contains two radiating components, each of which is made up of three elliptically shaped patches situated 60 degrees apart, and resembles the shape of a ﬂower. Moreover, the proposed antenna design incorporates a T-like ground branch that functions as a decoupling structure, and is composed of two modiﬁed inverted-L branches and an I-shaped stub, offering an isolation of more than 20 dB over the whole operation band (4.3–15.63 GHz). Furthermore, the proposed antenna system was fabricated and tested, and the envelope correlation coefﬁcient (ECC), diversity gain (DG), and total active reﬂection coefﬁcient (TARC), as well as the radiation characteristics and MIMO performance, were analyzed. The proposed UWB-MIMO antenna may be a suitable candidate for diverse UWB applications, based on the simulated and measured results of this study. The lengths of the modiﬁed T-like branches L g and L f have a dramatic impact on the UWB-MIMO system’s impedance-matching and isolation performance. Only the impact of these particular factors on system performance is examined, while other parameters are kept constant. The S-parameters for tuning L g from 12 mm to 14 mm are illustrated in Figure 6. The impedance bandwidth is further improved when L g changes from 12 mm to 14 mm, as shown in Figure 6a. Neither UWB features can be achieved when L g has a value of 12 mm or 13 mm. Different values of L g also exhibit various isolation effects in terms of their inﬂuence on isolation. Although the isolation effect is optimal overall when L g is at 14 mm, as shown in Figure 6b, the transmission coefﬁcient must be enhanced in the 6–8 GHz band region when L g is at 14 mm. These results demonstrate that the S-parameters of the system are signiﬁcantly inﬂuenced by the vertical length of the modiﬁed L-shaped ground branch (L g ), and it is best to select a value of 14 mm for L g when taking into account the size of the system. some frequencies with a difference of more than 20 dB can be seen. Manufacturing and measurement errors might be responsible for the discrepancy between the simulated and measured S12 results, which does not affect the overall high-isolation performance of the system. From these ﬁndings, the proposed antenna offers a wide operation band with high isolation.


Introduction
Ultra-wideband (UWB) technology has been extensively applied in areas including short-range communications, radar, location, and tracking due to its extraordinarily low transmission power and high data speed [1]. The employment of multiple-input, multipleoutput (MIMO) technology in a UWB wireless communication system enhances the signalto-noise ratio and data capacity of the communication system by enabling multiplexing, hence improving the overall performance of the system [2]. The performance of a UWB MIMO system is influenced by factors such as bandwidth, isolation, and other functional characteristics. In recent years, researchers have used a variety of ways to improve the performance of UWB antennas for MIMO systems.
In [3], an antenna designed on a 34 × 34 × 1.6 mm 3 FR4 substrate was proposed, with L-shaped and C-shaped slots embedded in each radiator and an electromagnetic bandgap (EBG) structure loaded near the microstrip feeding line, thereby allowing the antenna to exhibit triple-band slot characteristics, resulting in a bandwidth of 2.5-12 GHz. Here, in addition to the EBG structure being used to enhance the gain of the UWB antennas, the frequency selection surface (FSS) method will also have a significant gain-enhancing effect. Reference [4] presented a method for enhancing UWB antennas with FSS, which minimized power loss in the undesired transmission area of the antenna and blocked possible interference from undesirable and wasteful radiation, in order to obtain constant gain. Figure 1 depicts the topology of the proposed dual-port, flower-shaped UWB-MIMO antenna system, and Figure 2 shows the fabricated prototype. Compared with the antennas reported in [3,6,8,9,11,12,14], the MIMO antenna system proposed in this study has a smaller size of 30 × 18 mm 2 (0.84λ × 0.50λ), and was designed on an FR4 substrate with 1.6 mm thickness (tanδ = 0.02 and εr = 4.4). Two similar flower-shaped radiating elements and a metal ground make up the overall antenna model. Each flower-shaped radiating element directly supplied by a microstrip line is made up of three elliptically shaped patches set above the substrate, each at an angle of 60 degrees from the others. Next, improved and inverted L-shaped branches with mirror symmetry and an I-shaped stub above L-shaped branches are added to create a T-like branch at the bottom of the substrate and above the rectangular floor, thereby establishing the proposed ground structure with a rectangular floor. The role of the T-like branching in this design is comparable to that of the ladder resonator proposed in [16], which will effectively block or absorb the surface current between the patch antenna elements at the operating frequency, thereby reducing the mutual influence. The specific design process and principle are described below. The parameters of the proposed dual-port UWB-MIMO antenna are listed in Table 1.

Antenna Geometry
Electronics 2022, 11, x FOR PEER REVIEW 3 of 13 Figure 1 depicts the topology of the proposed dual-port, flower-shaped UWB-MIMO antenna system, and Figure 2 shows the fabricated prototype. Compared with the antennas reported in [3,6,8,9,11,12,14], the MIMO antenna system proposed in this study has a smaller size of 30 × 18 mm 2 (0.84λ × 0.50λ), and was designed on an FR4 substrate with 1.6 mm thickness (tanδ = 0.02 and εr = 4.4). Two similar flower-shaped radiating elements and a metal ground make up the overall antenna model. Each flower-shaped radiating element directly supplied by a microstrip line is made up of three elliptically shaped patches set above the substrate, each at an angle of 60 degrees from the others. Next, improved and inverted L-shaped branches with mirror symmetry and an I-shaped stub above Lshaped branches are added to create a T-like branch at the bottom of the substrate and above the rectangular floor, thereby establishing the proposed ground structure with a rectangular floor. The role of the T-like branching in this design is comparable to that of the ladder resonator proposed in [16], which will effectively block or absorb the surface current between the patch antenna elements at the operating frequency, thereby reducing the mutual influence. The specific design process and principle are described below. The parameters of the proposed dual-port UWB-MIMO antenna are listed in Table 1.   Figure 1 depicts the topology of the proposed dual-port, flower-shaped UWB-MIMO antenna system, and Figure 2 shows the fabricated prototype. Compared with the antennas reported in [3,6,8,9,11,12,14], the MIMO antenna system proposed in this study has a smaller size of 30 × 18 mm 2 (0.84λ × 0.50λ), and was designed on an FR4 substrate with 1.6 mm thickness (tanδ = 0.02 and εr = 4.4). Two similar flower-shaped radiating elements and a metal ground make up the overall antenna model. Each flower-shaped radiating element directly supplied by a microstrip line is made up of three elliptically shaped patches set above the substrate, each at an angle of 60 degrees from the others. Next, improved and inverted L-shaped branches with mirror symmetry and an I-shaped stub above Lshaped branches are added to create a T-like branch at the bottom of the substrate and above the rectangular floor, thereby establishing the proposed ground structure with a rectangular floor. The role of the T-like branching in this design is comparable to that of the ladder resonator proposed in [16], which will effectively block or absorb the surface current between the patch antenna elements at the operating frequency, thereby reducing the mutual influence. The specific design process and principle are described below. The parameters of the proposed dual-port UWB-MIMO antenna are listed in Table 1.

Design Evolution Stages of the MIMO Antenna
To examine the implications of different MIMO antenna configurations, the 50 Ω transmission line feed is utilized in combination with the fractional ground plane. The overall design procedure for the proposed UWB-MIMO antenna system is elaborated in Figure 3, and the MIMO system's working principle, using its reflection coefficient and transmission coefficient curves, is presented in Figures 4 and 5.

Design Evolution Stages of the MIMO Antenna
To examine the implications of different MIMO antenna configurations, the 50 Ω transmission line feed is utilized in combination with the fractional ground plane. The overall design procedure for the proposed UWB-MIMO antenna system is elaborated in Figure 3, and the MIMO system's working principle, using its reflection coefficient and transmission coefficient curves, is presented in Figures 4 and 5.    The radiating element in step 1 (Figure 3a) is made up of two mutually perpendi elliptical patches and a microstrip line, along with a full rectangular ground at the bo Notably, the radiating element in this case is identical to the one proposed in [17], w is utilized for 5G communication. Although the antenna designed in step1 can cover 12.85 GHz, its reflection coefficient is poor, and the best value for S11 is only −15.2 d shown in Figure 4. Meanwhile, the isolation between the antennas in the covered quency range is less than 19 dB, since there is no decoupling structure involved, as trated in Figure 5.
Therefore, the radiating element is modified by merging three elliptically sh patches in step 2, resembling a flower, but the ground structure is left unchanged. antenna structure generates two resonant modes at 8.4 GHz and 12.8 GHz from its r tion coefficient, and the impedance-matching performance is improved. In the reso modes, reflection coefficients are −43 and -21 dB, respectively, and the impedance b width reaches 4.89-14.13 GHz. The transmission coefficient between the antennas, o other hand, has not improved.
Progressively, in step 3, a horizontal I-shaped stub is placed above the invert shaped branch to produce a T-like branch on ground, which improves the imped  The radiating element in step 1 (Figure 3a) is made up of two mutually perpendicular elliptical patches and a microstrip line, along with a full rectangular ground at the bottom. Notably, the radiating element in this case is identical to the one proposed in [17], which is utilized for 5G communication. Although the antenna designed in step1 can cover 4.58-12.85 GHz, its reflection coefficient is poor, and the best value for S11 is only −15.2 dB, as shown in Figure 4. Meanwhile, the isolation between the antennas in the covered frequency range is less than 19 dB, since there is no decoupling structure involved, as illustrated in Figure 5.
Therefore, the radiating element is modified by merging three elliptically shaped patches in step 2, resembling a flower, but the ground structure is left unchanged. This antenna structure generates two resonant modes at 8.4 GHz and 12.8 GHz from its reflection coefficient, and the impedance-matching performance is improved. In the resonant modes, reflection coefficients are −43 and −21 dB, respectively, and the impedance bandwidth reaches 4.89-14.13 GHz. The transmission coefficient between the antennas, on the other hand, has not improved.
Progressively, in step 3, a horizontal I-shaped stub is placed above the inverted Lshaped branch to produce a T-like branch on ground, which improves the impedance matching and isolation compared with step 2. The antenna is stimulated into five resonant modes (5.38, 5.8, 7.8, 9.6, and 13 GHz), as shown by its reflection coefficient results, thus suggesting that the new T-like branch functions as a resonator, hence extending the bandwidth so that it spans between 4.48 and 15.26 GHz. From these results, we can see that the antenna's bandwidth has been increased. Furthermore, the total reflection coefficient is lowered, indicating that the impedance-matching ability is improved. Due to the separation impact of the T-like branch on the antenna components, isolation is enhanced, reaching more than 14.2 dB. These findings suggest that the adoption of the T-like branch is important for boosting the bandwidth and isolation.
In the MIMO antenna system, increasing the independence between the antenna components has long been a desired aim. However, as a result, additional reductions in the correlation and improved isolation between the antenna components are required. Accordingly, we enhanced the shape of the metallic ground in step 4, which completed the design of the proposed UWB-MIMO system. To construct the final ground structure, the shape of the inverted L-shaped branch was slightly modified, i.e., the vertical width was extended and the triangular patches were proportionately cut out at the edges of the inverted L-shaped branches. The resonance frequencies stimulated in step 4 are shifted to the right compared with those of step 3, which are now 5.4, 6, 8, 11.2, and 14.6 GHz, respectively, as displayed in Figure 4. Although the reflection coefficient performance is worse than that of step 3, the design of step 4 still covers the frequency range of 4.3-15.63 GHz. More importantly, Figure 5 shows that the isolation in step 4 has been enhanced compared with that in step 3, reaching more than 20 dB, which suggests that the mutual coupling has been decreased. Essentially, these results imply that the improved ground structure makes a significant contribution toward improving the isolation.

Parameter Analysis
The lengths of the modified T-like branches L g and L f have a dramatic impact on the UWB-MIMO system's impedance-matching and isolation performance. Only the impact of these particular factors on system performance is examined, while other parameters are kept constant. The S-parameters for tuning L g from 12 mm to 14 mm are illustrated in Figure 6. The impedance bandwidth is further improved when L g changes from 12 mm to 14 mm, as shown in Figure 6a. Neither UWB features can be achieved when L g has a value of 12 mm or 13 mm. Different values of L g also exhibit various isolation effects in terms of their influence on isolation. Although the isolation effect is optimal overall when L g is at 14 mm, as shown in Figure 6b, the transmission coefficient must be enhanced in the 6-8 GHz band region when L g is at 14 mm. These results demonstrate that the S-parameters of the system are significantly influenced by the vertical length of the modified L-shaped ground branch (L g ), and it is best to select a value of 14 mm for L g when taking into account the size of the system. lowered, indicating that the impedance-matching ability is improved. Due to the separation impact of the T-like branch on the antenna components, isolation is enhanced, reaching more than 14.2 dB. These findings suggest that the adoption of the T-like branch is important for boosting the bandwidth and isolation.
In the MIMO antenna system, increasing the independence between the antenna components has long been a desired aim. However, as a result, additional reductions in the correlation and improved isolation between the antenna components are required. Accordingly, we enhanced the shape of the metallic ground in step 4, which completed the design of the proposed UWB-MIMO system. To construct the final ground structure, the shape of the inverted L-shaped branch was slightly modified, i.e., the vertical width was extended and the triangular patches were proportionately cut out at the edges of the inverted L-shaped branches. The resonance frequencies stimulated in step 4 are shifted to the right compared with those of step 3, which are now 5.4, 6, 8, 11.2, and 14.6 GHz, respectively, as displayed in Figure 4. Although the reflection coefficient performance is worse than that of step 3, the design of step 4 still covers the frequency range of 4.3-15.63 GHz. More importantly, Figure 5 shows that the isolation in step 4 has been enhanced compared with that in step 3, reaching more than 20 dB, which suggests that the mutual coupling has been decreased. Essentially, these results imply that the improved ground structure makes a significant contribution toward improving the isolation.

Parameter Analysis
The lengths of the modified T-like branches Lg and Lf have a dramatic impact on the UWB-MIMO system's impedance-matching and isolation performance. Only the impact of these particular factors on system performance is examined, while other parameters are kept constant. The S-parameters for tuning Lg from 12 mm to 14 mm are illustrated in Figure 6. The impedance bandwidth is further improved when Lg changes from 12 mm to 14 mm, as shown in Figure 6a. Neither UWB features can be achieved when Lg has a value of 12 mm or 13 mm. Different values of Lg also exhibit various isolation effects in terms of their influence on isolation. Although the isolation effect is optimal overall when Lg is at 14 mm, as shown in Figure 6b, the transmission coefficient must be enhanced in the 6-8 GHz band region when Lg is at 14 mm. These results demonstrate that the S-parameters of the system are significantly influenced by the vertical length of the modified L-shaped ground branch (Lg), and it is best to select a value of 14 mm for Lg when taking into account the size of the system. The simulated variation in characteristics of Lf from 9 mm to 11 mm are presented in Figure 7. As can be seen from Figure 7a, the effect of Lf on return loss is not very significant at these values, and all can achieve acceptable ultra-wideband properties. However, the effect of Lf on isolation is even more pronounced. As can be observed in Figure 7b, the optimal isolation effect is obtained when Lf is 10 mm, that is, when the I-shaped ground The simulated variation in characteristics of L f from 9 mm to 11 mm are presented in Figure 7. As can be seen from Figure 7a, the effect of L f on return loss is not very significant at these values, and all can achieve acceptable ultra-wideband properties. However, the effect of L f on isolation is even more pronounced. As can be observed in Figure 7b, the optimal isolation effect is obtained when L f is 10 mm, that is, when the I-shaped ground stub and modified L-shaped ground branch are exactly combined to form a T-like branch, which also confirms the authenticity of the proposed decoupling structure.
Electronics 2022, 11, x FOR PEER REVIEW 7 of 13 stub and modified L-shaped ground branch are exactly combined to form a T-like branch, which also confirms the authenticity of the proposed decoupling structure.   Figure 8 presents the surface current distribution in the resonance modes, to visually emphasize the decoupling effect of ground plane geometry. In the proposed MIMO system, port 1 is stimulated, while port 2 is terminated with a 50 Ω matched load. When just port 1 is stimulated, the current is largely distributed on antenna 1 and its modified T-like branch on the same side, as shown in Figure 8, whereas the current distribution on the surface of antenna 2 is relatively weak. It can be deduced that current-absorbing effect of the improved ground branch successfully improves the port isolation between the two monopole antennas.

S-Parameter Results
HFSS and an Agilent N5247A vector network analyzer were used to simulate and measure the proposed UWB-MIMO antenna, and the results are provided in Figure 9. The measured S11 can cover 4.51-15.1 GHz, while S12 is below −15 dB, as shown in Figure 9. It is worthwhile to note that there are significant differences between the measured and  Figure 8 presents the surface current distribution in the resonance modes, to visually emphasize the decoupling effect of ground plane geometry. In the proposed MIMO system, port 1 is stimulated, while port 2 is terminated with a 50 Ω matched load. When just port 1 is stimulated, the current is largely distributed on antenna 1 and its modified T-like branch on the same side, as shown in Figure 8, whereas the current distribution on the surface of antenna 2 is relatively weak. It can be deduced that current-absorbing effect of the improved ground branch successfully improves the port isolation between the two monopole antennas.

Current Distribution
Electronics 2022, 11, x FOR PEER REVIEW 7 of 13 stub and modified L-shaped ground branch are exactly combined to form a T-like branch, which also confirms the authenticity of the proposed decoupling structure.   Figure 8 presents the surface current distribution in the resonance modes, to visually emphasize the decoupling effect of ground plane geometry. In the proposed MIMO system, port 1 is stimulated, while port 2 is terminated with a 50 Ω matched load. When just port 1 is stimulated, the current is largely distributed on antenna 1 and its modified T-like branch on the same side, as shown in Figure 8, whereas the current distribution on the surface of antenna 2 is relatively weak. It can be deduced that current-absorbing effect of the improved ground branch successfully improves the port isolation between the two monopole antennas.

S-Parameter Results
HFSS and an Agilent N5247A vector network analyzer were used to simulate and measure the proposed UWB-MIMO antenna, and the results are provided in Figure 9. The measured S11 can cover 4.51-15.1 GHz, while S12 is below −15 dB, as shown in Figure 9. It is worthwhile to note that there are significant differences between the measured and simulated results of S12, especially in 6.5-7.5 GHz and 10-11 GHz bands, where some

S-Parameter Results
HFSS and an Agilent N5247A vector network analyzer were used to simulate and measure the proposed UWB-MIMO antenna, and the results are provided in Figure 9. The measured S11 can cover 4.51-15.1 GHz, while S12 is below −15 dB, as shown in Figure 9. It is worthwhile to note that there are significant differences between the measured and simulated results of S12, especially in 6.5-7.5 GHz and 10-11 GHz bands, where some frequencies with a difference of more than 20 dB can be seen. Manufacturing and measurement errors might be responsible for the discrepancy between the simulated and measured S12 results, which does not affect the overall high-isolation performance of the system. From these findings, the proposed antenna offers a wide operation band with high isolation. s 2022, 11, x FOR PEER REVIEW frequencies with a difference of more than 20 dB can be seen. Manufactu urement errors might be responsible for the discrepancy between the simul ured S12 results, which does not affect the overall high-isolation performa tem. From these findings, the proposed antenna offers a wide operation b isolation.

Far-Field Properties
The UWB-MIMO antenna was tested in an anechoic chamber for its terns, and the results are presented in Figure 10. By stimulating port 1 a port 2 with a matched load, the corresponding radiation patterns were m be seen from Figure 10a,b that the antenna almost achieved omnidirection both the XOZ plane and the YOZ plane at low frequencies (5.4 GHz and 6 G from (c), the radiation direction of the antenna on the XOZ surface was mai between 180° and 360°, and the maximum radiation gain reached 4.4 dB at 255°, while on the YOZ surface, the maximum radiation direction was ap and 180°. At 11.2 GHz, the XOZ surface achieved almost omnidirectional acteristics, while on the YOZ surface, the radiation gain obtained its max approximately 90°. However, at high frequency, it can be seen from Figu radiation pattern of the XOZ plane was not as good as that of the YOZ pl to these results, the antenna had acceptable radiation characteristics on the and YOZ (H-plane) at frequencies of 5.4, 6, 8, 11.2, and 14.6 GHz, and the m are consistent with the simulated results.

Far-Field Properties
The UWB-MIMO antenna was tested in an anechoic chamber for its radiation patterns, and the results are presented in Figure 10. By stimulating port 1 and terminating port 2 with a matched load, the corresponding radiation patterns were measured. It can be seen from Figure 10a,b that the antenna almost achieved omnidirectional radiation on both the XOZ plane and the YOZ plane at low frequencies (5.4 GHz and 6 GHz). At 8 GHz from (c), the radiation direction of the antenna on the XOZ surface was mainly distributed between 180 • and 360 • , and the maximum radiation gain reached 4.4 dB at approximately 255 • , while on the YOZ surface, the maximum radiation direction was approximately 0 • and 180 • . At 11.2 GHz, the XOZ surface achieved almost omnidirectional radiation characteristics, while on the YOZ surface, the radiation gain obtained its maximum value at approximately 90 • . However, at high frequency, it can be seen from Figure 10e that the radiation pattern of the XOZ plane was not as good as that of the YOZ plane. According to these results, the antenna had acceptable radiation characteristics on the XOZ (E-plane) and YOZ (H-plane) at frequencies of 5.4, 6, 8, 11.2, and 14.6 GHz, and the measured results are consistent with the simulated results.
The proposed MIMO antenna's radiation efficiency and peak gain are shown on Figure 11. As the radiators of the proposed MIMO antenna were structured symmetrically, the following values are shown for only one radiator. The radiation efficiency varied from 85 to 93 percent, implying that the majority of the energy was radiated away. The maximum gain of a single antenna at 5.4 GHz was 5.35 dBi, whereas the peak gain values at 6.4-7.6 GHz were relatively low, in the range of 2.5-4 dBi. Throughout the whole operational frequency range, the proposed MIMO antenna showed a positive gain value. Based on the results described above, the proposed UWB-MIMO antenna system offers promising radiation features. acteristics, while on the YOZ surface, the radiation gain obtained its maximum value at approximately 90°. However, at high frequency, it can be seen from Figure 10e that the radiation pattern of the XOZ plane was not as good as that of the YOZ plane. According to these results, the antenna had acceptable radiation characteristics on the XOZ (E-plane) and YOZ (H-plane) at frequencies of 5.4, 6, 8, 11.2, and 14.6 GHz, and the measured results are consistent with the simulated results. The proposed MIMO antenna's radiation efficiency and peak gain are shown on Figure 11. As the radiators of the proposed MIMO antenna were structured symmetrically, the following values are shown for only one radiator. The radiation efficiency varied from 85 to 93 percent, implying that the majority of the energy was radiated away. The maximum gain of a single antenna at 5.4 GHz was 5.35 dBi, whereas the peak gain values at 6.4-7.6 GHz were relatively low, in the range of 2.5-4 dBi. Throughout the whole operational frequency range, the proposed MIMO antenna showed a positive gain value. Based on the results described above, the proposed UWB-MIMO antenna system offers promising radiation features.  The proposed MIMO antenna's radiation efficiency and peak gain are shown on ure 11. As the radiators of the proposed MIMO antenna were structured symmetric the following values are shown for only one radiator. The radiation efficiency varied 85 to 93 percent, implying that the majority of the energy was radiated away. The m mum gain of a single antenna at 5.4 GHz was 5.35 dBi, whereas the peak gain valu 6.4-7.6 GHz were relatively low, in the range of 2.5-4 dBi. Throughout the whole op tional frequency range, the proposed MIMO antenna showed a positive gain value. B on the results described above, the proposed UWB-MIMO antenna system offers pro ing radiation features.  Figure 11. Calculated radiation efficiency and peak gain.

MIMO Performance
Different performance metrics of MIMO antennas, such as ECC (envelope correla coefficient), DG (diversity gain), and TARC (total active reflection coefficient), shoul examined to ensure their efficient operation.
The ECC value is a critical parameter for evaluating the performance of the radia patterns of MIMO radiators. A lower ECC value indicates a lower effect on other ante while working alone, and greater efficiency. To guarantee effective operation of each tenna, the specified ECC value of a MIMO system in wireless communication networ generally less than 0.5 [18]. The ECC can be calculated from the S-parameters using E tion (1). The results of simulated and measured ECC are displayed in Figure 12, w shows an acceptable isolation performance.

MIMO Performance
Different performance metrics of MIMO antennas, such as ECC (envelope correlation coefficient), DG (diversity gain), and TARC (total active reflection coefficient), should be examined to ensure their efficient operation.
The ECC value is a critical parameter for evaluating the performance of the radiation patterns of MIMO radiators. A lower ECC value indicates a lower effect on other antennas while working alone, and greater efficiency. To guarantee effective operation of each antenna, the specified ECC value of a MIMO system in wireless communication networks is generally less than 0.5 [18]. The ECC can be calculated from the S-parameters using Equation (1). The results of simulated and measured ECC are displayed in Figure 12, which shows an acceptable isolation performance. ECC can be computed from the S-parameters using the following equation.
Another key MIMO performance metric is DG, which describes how effective the diversity is. ECC is used to represent its value, which can be computed by Equation (2).
The value of DG, calculated from the S-parameters of the proposed MIMO antenna system, is shown in Figure 13. In the working frequency band, DG is larger than 9.96 dB, thereby indicating a strong MIMO diversity performance.  ECC can be computed from the S-parameters using the following equation.
Another key MIMO performance metric is DG, which describes how effective the diversity is. ECC is used to represent its value, which can be computed by Equation (2).
The value of DG, calculated from the S-parameters of the proposed MIMO antenna system, is shown in Figure 13. In the working frequency band, DG is larger than 9.96 dB, thereby indicating a strong MIMO diversity performance. ECC can be computed from the S-parameters using the following equation.
Another key MIMO performance metric is DG, which describes how effective the diversity is. ECC is used to represent its value, which can be computed by Equation (2).
The value of DG, calculated from the S-parameters of the proposed MIMO antenna system, is shown in Figure 13. In the working frequency band, DG is larger than 9.96 dB, thereby indicating a strong MIMO diversity performance. Furthermore, the statistic TARC is related to the total reflected power and total incident power, and it is used to assess the MIMO system's effectiveness. TARC should ideally be zero, which indicates that the antenna receives all of the incident power [19]. TARC Furthermore, the statistic TARC is related to the total reflected power and total incident power, and it is used to assess the MIMO system's effectiveness. TARC should ideally be zero, which indicates that the antenna receives all of the incident power [19]. TARC can be computed using Equation (3) for a two-port system. Figure 14 displays a comparison between simulated and measured TARC values, which reveals that TARC is less than −20 dB in operating frequency range.
can be computed using Equation (3) for a two-port system. Figure 14 displays a comparison between simulated and measured TARC values, which reveals that TARC is less than −20 dB in operating frequency range.  Table 2 compares the performance of the proposed structure with the performance of various previously published UWB-MIMO antenna configurations. In contrast with the other reported designs, the proposed UWB-MIMO antenna structure provides obvious benefits. As can be observed from the table, the proposed UWB-MIMO antenna structure outperforms all the other designs. The proposed candidate is ideally suited for numerous UWB wireless applications owing to its small size, moderate impedance bandwidth, strong isolation capability, low ECC, high gain, and relatively steady radiation efficiency.   Table 2 compares the performance of the proposed structure with the performance of various previously published UWB-MIMO antenna configurations. In contrast with the other reported designs, the proposed UWB-MIMO antenna structure provides obvious benefits. As can be observed from the table, the proposed UWB-MIMO antenna structure outperforms all the other designs. The proposed candidate is ideally suited for numerous UWB wireless applications owing to its small size, moderate impedance bandwidth, strong isolation capability, low ECC, high gain, and relatively steady radiation efficiency.

Conclusions
A compact-sized, dual-port, flower-shaped UWB-MIMO antenna with high isolation is presented in this paper. Three elliptically shaped metal patches, located at 60 degrees from each other, were combined to form a flower-shaped radiating element. The adopted ground structure improvements, including two modified inverted L-shaped branches and an I-shaped stub, expand the impedance bandwidth by generating multiple resonance modes to cover the 4.3-15.63 GHz (relative bandwidth 113.4%) range, which is typically used for multi-standard wireless applications, such as 5G N79 (4.4-5 GHz), WLAN (5.15-5.35 GHz/5.72-5.825 GHz), 5G spectrum band (5.9-6.4 GHz), X-band for satellite communication (8)(9)(10)(11)(12), FSS (11.45-11.7 GHz/12.5-12.75 GHz), and Ku band (12)(13)(14)(15)(16)(17)(18). These improvements also effectively reduce the mutual coupling between the antennas by absorbing the current, thereby enhancing the isolation to more than 20 dB. In addition, the measured results suggest that the proposed antenna displays favorable radiation patterns, where the radiation efficiency is between 85% and 93%, while the peak gain ranges from 2.  Data Availability Statement: The simulated and measured data used to support the findings of this study are included within the article.

Conflicts of Interest:
The authors declare no conflict of interest.