1 Introduction

Nowadays, mobile communication has taken its place as an indispensable part of our daily lives. Especially in the last decades, mobile communication has made tremendous progress with the growing demand for mobility, and its evolution from the first generation (1G) to the fifth generation (5G) represents a significant milestone in wireless communication [1]. As current wireless communication technologies primarily utilize sub-6 GHz bands, there is a pressing need for additional bandwidth and capacity to support the increasing data traffic. This has led to the exploration of more efficient utilization of the available spectrum, particularly at high frequencies such as millimeter-wave (mm-wave) [2]. Although mm-waves face challenges like high atmospheric absorption losses and production costs, several frequency bands for 5G mm-wave applications have been defined by ITU. These include the 24.25–27.5 GHz (n258), 26.5–29.5 GHz (n257), 27.5–28.35 GHz (n261), 37–40 GHz (n260), and 47.2–48.2 GHz (n262) bands [3]-23[4]. Among these bands, the n257 (28 GHz) band has gained significant attention in both industry and academia due to its lower atmospheric attenuation and global availability [1].

Printed antennas are particularly suitable for 5G mm-wave communication due to their benefits, such as ease of manufacturing, compact size, low cost, and compatibility with Monolithic Microwave Integrated Circuits (MMIC). However, these antennas also possess certain drawbacks, including narrow bandwidth and low gain [5]. In the case of mm-wave antennas, which require wide bandwidth, high gain, and efficiency, these limitations pose significant challenges due to factors such as high atmospheric absorption or limited radiant power capacity at high frequencies. Various methods have been proposed in the literature to address these issues, including the use of high relative permeability substrates, incorporation of slots or shorting pins, utilization of defected ground structures (DGS), addition of parasitic elements, adoption of frequency-independent antenna architectures like log-periodic structures, and using of metamaterial-based techniques such as superstrates, frequency selective surfaces (FSS), and electromagnetic band gaps (EBG) [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19].

Furthermore, multi-port antenna technology has gained widespread adoption in mm-wave communication due to its ability to increase data transfer rates and ensure stable communication quality. The miniaturization of mobile devices necessitates the integration of multi-port antennas within a limited space. Designing multi-port antennas that maintain high isolation levels, low envelope correlation coefficients (ECC), and do not require decoupling structures is critical [20,21,22,23].

Over the past few years, there have been numerous advancements in the development of multi-port 5G antenna designs for the 28 GHz mm-wave [24,25,26,27,28,29]. For instance, one such design presented in [24] introduces a four-port MIMO antenna with T-shaped radiating elements. These T-shaped radiators are fed by coplanar waveguides (CPW), and split ring slots are incorporated into the defected ground planes. The proposed four-port MIMO antenna, etched on a Rogers RT-5880 substrate measuring 50.8 × 12 mm2, offers a frequency bandwidth of 25.1–37.5 GHz and achieves a maximum gain of 10.6 dBi at 36 GHz. Another MIMO configuration described in [25] features a dielectric resonator antenna (DRA) with overall dimensions of 20 × 20 × 2.8 mm3. This design covers a band of 0.85 GHz within the 27.5–28.35 GHz frequency range, providing a maximum isolation level of 24 dB and a peak gain of 8 dBi. In [26], a dual-port 2 × 8 MIMO antenna array with a substrate size of 23.6 × 55.1 mm2 is proposed, offering an operating bandwidth of 27.49–29.42 GHz. This design achieves a peak gain of 11.33 dBi. Compact-sized MIMO antennas are also explored in [27] and [28], demonstrating good isolation levels and maximum gain values of 5.4 dBi and 5.9 dBi, respectively. Lastly, [29] proposes a four-port MIMO antenna with infinity shell-shaped patches featuring a substrate dimension of 30 × 30 mm2. This design offers approximately 3.9 GHz bandwidth, an isolation level of 29 dB, and a peak gain of 6.1 dBi. These advancements in multi-port antenna designs highlight the ongoing efforts to optimize performance metrics such as bandwidth, isolation, and gain for 5G n257 band applications.

This paper proposes an improved dual-port PMLPMA (Printed Meandered Log-Periodic Monopole Array) antenna system for 28 GHz 5G mobile applications. The proposed antenna system is designed to achieve enhanced gain performance while maintaining a compact size. A novel octagonal-ring FSS layer is integrated into the antenna system to achieve gain improvement. The simulated results show that the dual-port antenna system with the FSS layer exhibits a bandwidth ranging from 26.5 to 29.5 GHz. The simulated and measured results demonstrate isolation levels exceeding 30 dB across the operational band. Moreover, the addition of the octagonal FSS layer results in a significant gain improvement of over 65% throughout the entire operating band of the antenna system. The proposed antenna system features a unique mid-point curved log-periodic architecture, distinguishing it from previous studies in the literature. This architecture not only provides the advantages of wideband performance associated with log-periodic patch antennas but also reduces the overall size of the antenna.

2 Design Procedures for the Dual-Port PMLPMA Antenna System

The geometry of the proposed PMLPMA and its dual-port configuration are presented in this part of the paper. Furthermore, parametric studies are also analyzed in this section. All the simulations are carried out by CST Studio Suite software.

2.1 Design Methodology of the Proposed PMLPMA

The proposed PMLPMA structure is inspired by conventional PLPDA (Printed Log-Periodic Dipole Array), as depicted in Fig. 1 [30]. The performance of the proposed PMLPMA, as in conventional PLPDA, is widely affected by various design parameters like spacing factor (σ), scale factor (τ), longest dipole length (Lmax), feed length (Lf), and number of dipole elements (N).

Fig. 1
figure 1

Basic PLPDA architecture

Full size image

The design procedure for the proposed PMLPMA begins with determining the minimum and maximum cut-off frequencies (fmin and fmax) of the log-periodic array. Subsequently, the important design parameters of the dielectric constant or relative permittivity (εr) and substrate height (hs) that impact the antenna radiation performance are identified. Then, effective relative permittivity (εeff) is obtained as follows:

$$\varepsilon_{eff} = \frac{{\varepsilon_{r} + 1}}{2} + \frac{{\varepsilon_{r} - 1}}{2}\frac{1}{{\sqrt {1 + 12h_{s} /w} }}$$
(1)

where the width of the microstrip-fed (wf) is calculated as 1.55 mm by simultaneously solving Eqs. (1) and (2).

$$Z_{c} = \frac{120\pi }{{\sqrt {\varepsilon_{eff} } \left[ {\frac{w}{{h_{s} }} + 1.393 + \frac{2}{3}\ln \left( {\frac{w}{{h_{s} }} + 1.444} \right)} \right]}}$$
(2)

Here Zc represents the microstrip line's characteristic impedance, typically assumed to be 50 Ω, which corresponds to the impedance of the standard SMA connector. Thereafter, the length of the feed line (Lf) is calculated as 3.9 mm by Eq. (3);

$$L_{f} = \frac{c}{{2f_{0} \sqrt {\varepsilon_{eff} } }}$$
(3)

where c denotes the speed of light in free space. The resonant frequency of the antenna (f0) is set to 28 GHz for this study. Afterward, the length of the longest log-periodic element (Lmax) is determined to be approximately 2.07 mm using Eq. (4).

$$L_{\max } = L_{1} = \frac{{\lambda_{\max } }}{{4\sqrt {\varepsilon_{eff} } }} = \frac{c}{{4f_{\min } \sqrt {\varepsilon_{eff} } }}$$
(4)

Then, the log-periodic antenna input parameters τ and σ need to be established. Equation (5) is used to calculate the space (Sn) between sequential array elements using the length of the nth element (Ln) and the space factor (σ).

$$\sigma = \frac{{s_{n} }}{{4L_{n} }}$$
(5)

The lengths of the log-periodic array elements are computed based on the scale factor (τ) using Eq. (6).

$$\tau = \frac{{L_{n + 1} }}{{L_{n} }} = \frac{{w_{n + 1} }}{{w_{n} }} = \frac{{s_{n + 1} }}{{s_{n} }}$$
(6)

The total number of log-periodic elements (N) can be determined by Eq. (7).

$$N = 1 + \frac{{\log (B.B_{ar} )}}{\log (1/\tau )}$$
(7)

where Bar and B are the active region bandwidth and expected bandwidth of the log-periodic array, respectively. Bar and B are obtained from Eqs. (8) and (9), while the half-aperture angle of the log-periodic array (α) is calculated using Eq. (10).

$$B_{ar} = 1.1 + 7.7(1 - \tau )^{2} \cot \alpha$$
(8)
$$B = \frac{{f_{\min } }}{{f_{\max } }}$$
(9)
$$\alpha = \tan^{ - 1} \left( {\frac{1 - \tau }{{4\sigma }}} \right)$$
(10)

From the aforementioned equations, τ is optimized as 0.78, and the accompanying spacing factor (σ) is set as 0.17 for the proposed PMLPMA Furthermore, the parameters α and N are calculated as 35.8° and 5, respectively.

2.2 PMLPMA Geometry

Figure 2 shows 2D geometric views of the unit PMLPMA, while Table 1 provides the detailed dimensions of the PMLPMA. The proposed PMLPMA is designed to operate at 28 GHz band with a compact size of 4.15 × 13 mm2. The proposed PMLPMA design is implemented on a double-sided Rogers RT-5880 dielectric laminate with a relative permittivity (εr) of 2.2 and a substrate height (hs) of 0.508 mm. Unlike conventional PLPDAs, the proposed PMLPMA distinguishes itself through its unique structure, employing monopole elements that resemble mid-point bent radiator elements. The radiator of the PMLPMA consists of five meandered log-periodic monopole elements, and it is fed by a 50 Ω microstrip line. Additionally, there is a thin impedance matching line between the PMLPMA patch and the microstrip feed line. On the backside of the laminate, a square ground plane with a side length of 4.15 mm is positioned to serve its purpose.

Fig. 2
figure 2

Proposed PMLPMA geometry a top view b bottom view c side view

Full size image
Table 1 The detailed dimensions of the unit PMLPMA structure (in mm)
Full size table

2.3 Dual-Port PMLPMA antenna

The proposed dual-port antenna system consists of two PMLPMA elements positioned orthogonally to each other, as illustrated in Fig. 3. The dual-port configuration of the PMLPMA antenna exhibits a highly compact structure, with dimensions of 18.75 × 18.75 mm2. Figure 4 presents a comparison of the scattering parameters (s-parameters) between the individual unit PMLPMA and its dual-port configuration. The simulation results demonstrate that both the unit PMLPMA and its dual-port configuration exhibit similar return loss (S11) characteristics. As observed in Fig. 4, the − 10 dB bandwidth of the dual-port PMLPMA spans from 26.6 to 30.1 GHz, while the unit PMLPMA antenna operates within a bandwidth ranging from 26.55 to 29.5 GHz. It is seen that both cases closely align with the target n257 band. Additionally, the proposed antenna achieves an isolation level exceeding 30 dB between the PMLPMA elements. Consequently, there is no requirement for any decoupling structure to mitigate the effects of mutual coupling.

Fig. 3
figure 3

2D architecture of the dual-port PMLPMA antenna

Full size image
Fig. 4
figure 4

Comparative s-parameters of the PMLPMA and its dual-port configuration

Full size image

To see the mutual coupling between the proposed dual-port antenna, surface current distributions were investigated separately for both ports at 28 GHz frequency using CST software, as shown in Fig. 5. While examining the current distribution, one port was excited and the other port was terminated with 50 Ω. It can be noticed that a small amount of current is generated in the terminated antenna as shown in the figure.

Fig. 5
figure 5

The simulated surface current distributions of the dual-port PMLPMA antenna

Full size image

3 Octagonal Ring FSS

One of the key steps in eliminating path losses is to enhance the antenna gain. A significant technique for improving gain involves incorporating metamaterial-based layers like EBG, superstrate, or FSS. Figure 6 illustrates the setup of the antenna coupled with an FSS layer. In the figure, hss and d correspond to the height of the superstrate layer and the gap between the two substrates, respectively. To optimize the antenna gain, it is recommended that these two variables be approximately one-fourth of the wavelength [31].

Fig. 6
figure 6

Side view of an antenna with an FSS layer

Full size image

The CST Studio Suite optimization tool was utilized to optimize the hss and d parameters for achieving a resonant frequency of 28 GHz. The resulting optimized values were 2.7 mm for hss and 1.55 mm for d. Figure 7 illustrates the proposed configuration of the octagonal ring FSS unit cell and its periodic structure. The unit cell is composed of two identical square rings that are interconnected perpendicular to each other. For the design of the unit FSS cell, Rogers RO3003 laminate was used, with a relative permittivity of 3, a height of 1.52, and a tangent loss of 0.0012.

Fig. 7
figure 7

The octagonal ring FSS a unit cell geometry b periodic structure

Full size image

The substrate length of the unit FSS cell (wsub) was determined to be 3.75 mm using the equation given in Eq. (11). Through optimization, the conductor length of the FSS patch (Lp) and the width of the octagonal ring (wp) were set to 3.5 and 0.1 mm, respectively. By referring to the figure, the spacing (s) between the FSS patch and the substrate edge was found to be 0.125 mm.

$$W_{sub} = \frac{c}{{2f_{0} \sqrt {\frac{{\varepsilon_{r} + 1}}{2}} }}$$
(11)

The proposed octagonal ring FSS has been analyzed with simulations in terms of transmission coefficient (S21) and reflection coefficient (S11) according to variation of the incidence angle (θ). Figure 8 depicts the incidence angular characteristics of the octagonal ring FSS, considering both the TE and TM planes. It can be observed from the graphs that when θ = 0°, S21 remains below − 10 dB within the 25.4–30.7 GHz range, with the resonance frequency occurring at 28 GHz. The graphs indicate a slight shift in the resonant frequency for different theta angles in both planes.

Fig. 8
figure 8

Transmission (S21) and reflection (S11) coefficients of the proposed FSS at various theta (θ) angles for: a TE plane b TM plane

Full size image

Following the FSS design process, simulations were made for the dual-port PMLPMA antenna both with and without the FSS. The obtained peak gain and total efficiency results are compared in Fig. 8. After the FSS design process, performance analyses of the proposed antenna system were made for both with/without FSS cases. Comparative plots of the peak gain, total efficiency, and s-parameters for the proposed antenna are presented in Fig. 9.

Fig. 9
figure 9

Comparative performance plots of the dual-port PMLPMA antenna system with/without the octagonal ring FSS: a peak gain and total efficiency b s-parameters

Full size image

As seen in Fig. 9a, the addition of a single layer FSS on top of the proposed dual-port PMLPMA antenna resulted in a notable increase in the maximum gain, rising from 3.5 dBi to approximately 7.35 dBi. Moreover, the dual-port antenna with FSS exhibited an overall efficiency improvement of approximately 5%. Gain-enhancing elements must have minimal impact on the operating band while increasing antenna performance. Therefore, the impact of the proposed octagonal-shaped FSS layer on the s-parameters of the antenna system was investigated and compared. The analysis revealed that in Fig. 9b, the proposed antenna with FSS exhibited an impedance bandwidth of 26.15–29.65 GHz, with a resonance frequency of 28.25 GHz. It can be concluded that the addition of FSS did not significantly alter the impedance bandwidth of the dual-port PMLPMA antenna. Only a slight shift in the resonant frequency was observed. Similarly, there is a slight difference in the isolation level between the antenna elements.

4 Parametric Studies

The parametric analysis of the dual-port PMLPMA antenna began with an investigation into the design parameters of the unit PMLPMA antenna (Fig. 10). In Fig. 10a, the impact of the ground plane width (Wg) and length (Lg) on the operating band and resonance frequency of the unit PMLPMA antenna is investigated. The graphs reveal that increasing Lg from 3.9 to 4.4 mm results in an expansion of the antenna bandwidth and an increase in return loss while shifting the resonance frequency towards a higher frequency. Conversely, as the Wg parameter increases, the resonance frequency decreases while the return loss value increases. Similarly, Fig. 10b illustrates the variation of the proposed PMLPMA bandwidth through varying matching line width (We) and length (Le). The figure demonstrates that enlarging both Le and We dimensions has a lowering effect on the resonance frequency. Furthermore, it is observed that the return loss reaches its maximum level when Le and We are set to 4.15 mm and 0.45 mm, respectively.

Fig. 10
figure 10

Effects of some design parameters on the return loss of the proposed PMLPMA a Lg and Wg b Le and We

Full size image

Additionally, the effects of the length of the unit FSS patch (Lp) and width of the FSS ring (wp) on the FSS transmission loss are also studied, as shown in Fig. 11. The variations in Lp size impact both the resonant frequency and the operating bandwidth of the unit FSS cell. Increasing the Lp from 3.3 to 3.7 mm led to a decrease in the resonant frequency from 30.3 to 25.6 GHz. Similarly, adjusting the size of the wp resulted in a comparable effect on the bandwidth and resonant frequency of the FSS. Increasing the wp caused the resonant frequency of the FSS to shift from 27.5 to 28.5 GHz.

Fig. 11
figure 11

Effects of the wp and Lp parameters on the transmission loss of the proposed FSS

Full size image

5 Results and Discussion

The dual-port PMLPMA antenna and octagonal ring FSS prototypes were fabricated on dielectric substrates of RT-5880 and RO3003, respectively. The dimensions of the substrates were 18.75 × 18.75 mm2, as shown in Fig. 12.

Fig. 12
figure 12

Photographs of prototyped elements: a PMLPMA MIMO antenna b octagonal ring FSS

Full size image

Following the fabrication process, performance measurements of the proposed MIMO antenna were made in a semi-anechoic chamber utilizing a Keysight PNA-X 5244B vector network analyzer and an EBTRO EAMS-120 turntable. The measurement setup inside the semi-anechoic chamber is given in Fig. 13. This setup is crucial for reliable measurements of the proposed antenna system's performance.

Fig. 13
figure 13

Experimental setup for antenna measurement

Full size image

Figure 14 shows the comparative s-parameter graphs for measured and simulated results of the proposed dual-port PMLPMA antenna with and without the octagonal ring FSS. The dual-port PMLPMA antenna without FSS is measured in the frequency range of 26.85 to 29.4 GHz, while the operating band of the antenna without FSS is measured as 26.65–29.75 GHz (Fig. 14a). The resonance frequencies for the antenna with and without FSS were also measured at 28.2 GHz and 27.7 GHz, respectively. Compared to the simulated results, the bandwidths of both the antenna system with and without FSS are largely similar, with a slight change in resonant frequencies. In addition, when the transmission coefficient (S21) curves were examined, it was determined that the measured results are much better than the simulated ones. Therefore, although there are some differences in terms of resonance frequency and operating bandwidth, the simulation and measurement results of the proposed antenna largely overlap with the determined design values and cover almost the entire 5G n257 band.

Fig. 14
figure 14

Measured and simulated s-parameters of the dual-port PMLPMA antenna: a S11 b S21

Full size image

Finally, the radiation pattern performance evaluations of the dual-port antenna are examined. In Fig. 15, we compare the measured and simulated polar radiation patterns of the proposed antenna, both with and without FSS, at a frequency of 28 GHz. Figure 15a presents the E-plane (xz-plane) pattern, while Fig. 15b depicts the H-plane (yz-plane) pattern of the antenna. The polar axis, spanning from 0° to 360°, represents the azimuth axis, defining angles in a circular pattern around the antenna. On the axis perpendicular to the polar axis, we observe the gain values corresponding to the azimuth angle.

Fig. 15
figure 15

Comparative radiation patterns of the dual-port PMLPMA antenna at 28 GHz a xz-plane; b yz-plane

Full size image

As can be seen from the figure, the radiation patterns of both dual-port antenna with and without FSS were quite similar to each other. At the 28 GHz frequency point, the maximum gain values for the dual-port PMLPMA antenna with FSS were measured as 6.8 dBi in the xz-plane and 5.95 dBi in the yz-plane. In contrast, the maximum gains measured in the antenna without FSS were 3.1 dBi and 2.1 dBi in the E and H planes, respectively.

5.1 The Dual-Port PMLPMA Antenna’s Diversity Performance

The diversity capability of the proposed antenna is verified through the examination of several performance attributes, namely the envelope correlation coefficient (ECC), diversity gain (DG), total active reflection coefficient (TARC), and channel capacity loss (CCL) [21].

ECC is a crucial parameter that quantifies the correlation level between multiple ports within an antenna system. In an ideal scenario, the ECC value should be precisely zero, indicating complete independence and isolation between the ports. However, in practical situations, an ECC value of less than 0.05 is considered adequate to achieve the best antenna performance. The ECC, denoted by the symbol ρe, can be computed using Eq. (12) by employing the radiation pattern of the antenna elements, as detailed in reference [20]. This measurement allows for a comprehensive assessment of the antenna's diversity capability, which is vital for enhancing its overall performance in wireless communication systems.

$$\rho_{e} = \frac{{\left| {\iint_{4\pi } {\left[ {\overrightarrow {{F_{1} }} (\theta ,\phi )*\overrightarrow {{F_{2} }} (\theta ,\phi )} \right]d\Omega }} \right|^{2} }}{{\iint_{4\pi } {\left| {\overrightarrow {{F_{1} }} (\theta ,\phi )} \right|^{2} d\Omega }\iint_{4\pi } {\left| {\overrightarrow {{F_{2} }} (\theta ,\phi )} \right|^{2} d\Omega }}}$$
(12)

In the computation of ECC, several parameters come into play. The symbol Ω represents the solid angle, while the * denotes the Hermitian product operator. Fi(θ,ϕ) displays the antenna's 3D field radiation pattern of the antenna when the ith port is activated. The ECC can also be determined using S-parameters, as illustrated in Eq. (13).

$$\rho_{e} = \frac{{\left| {S_{11}^{*} S_{12} + S_{21}^{*} S_{22} } \right|^{2} }}{{\left( {1 - \left| {S_{11} } \right|^{2} - \left| {S_{21} } \right|^{2} } \right)\left( {1 - \left| {S_{22} } \right|^{2} - \left| {S_{12} } \right|^{2} } \right)}}$$
(13)

The calculation of ECC involves complex mathematical operations and the examination of the antenna's radiation patterns and S-parameters. This comprehensive analysis helps in understanding the level of correlation between different ports in the antenna system, which in turn plays a critical role in optimizing the antenna's diversity capability and overall performance in practical wireless communication applications.

The received signals are combined using some diversity techniques to increase the signal-to-noise ratio (SNR) by using multiple antennas instead of a single antenna element. The diversity gain (DG) plays a significant role in characterizing this variation and is closely related to the correlation parameter ρe, as depicted in Eq. (14). The diversity gain provides valuable insights into the effectiveness of these techniques in mitigating signal fading and enhancing the overall communication performance.

$$DG = 10\sqrt {1 - \left( {\rho_{e} } \right)^{2} }$$
(14)

In Fig. 16a, a comparison between the measured and simulated ECC and DG curves of the dual-port PMLPMA antenna with an octagonal ring FSS layer is presented. The results show that the proposed dual-port PMLPMA antenna system exhibits a correlation well below 0.5, which meets the practical standard for ECC. Similarly, the DG remains stable at approximately 10 dB across the n257 band, indicating strong diversity performance of the antenna.

Fig. 16
figure 16

Compared diversity characteristics of the proposed PMLPMA antenna system with FSS. a ECCvsDG; b TARCvsCCL 

Full size image

The scattering parameters (s-parameters) are insufficient to characterize the efficiency and bandwidth (BW) of the proposed antenna system. The total active reflection coefficient (TARC) is used to better characterize the multi-port antenna system. The TARC can be calculated using the s-parameters of the antenna as given in Eq. (15). In a multi-port antenna system, it is desirable to be less than − 10 dB TARC value over the operating band for a good diversity performance.

$$TARC = \sqrt {\frac{{\left( {S_{ii} + S_{ij} } \right)^{2} + \left( {S_{ji} + S_{jj} } \right)^{2} }}{2}}$$
(15)

Channel capacity loss (CCL) is another key metric of diversity performance in multi-port antennas, and it indicates the maximum data rate at which transmission can continue with little loss. CCL can also be calculated using the s-parameters of the antenna. Equation (16) gives the formula for CCL. In a multi-port antenna system, the maximum allowed CCL value is less than 0.4 bits/s/Hz.

$$CCL = - \log_{2} \left\{ {\det (\psi^{R} )} \right\}$$
(16)

where ψR is the correlation matrix found by the equations given below.

$$\psi^{R} = \left[ {\begin{array}{*{20}c} {\sigma_{11} } & {\sigma_{12} } \\ {\sigma_{21} } & {\sigma_{22} } \\ \end{array} } \right]$$
(17)
$$\sigma_{ii} = 1 - \left( {\left| {S_{ii} } \right|^{2} - \left| {S_{ij} } \right|^{2} } \right)$$
(18)
$$\sigma_{ij} = - \left( {S_{ii} *S_{ij} + S_{ji} S_{ii} *} \right)$$
(19)

Figure 16b shows the TARC and CCL performances of the dual-port PMLPMA antenna. The graphs show that both the measured and simulated TARC values are consistently below − 10 dB throughout the operating band. Likewise, the graphs demonstrate that the CCL value is substantially lower than the threshold value, which indicates that the proposed antenna system is viable for 28 GHz 5G applications.

5.2 Comparison with Related Studies

Table 2 compares the proposed PMLPMA antenna system’s performance parameters with the state-of-the-art 28 GHz dual-port antennas presented in [25,26,27,28]. Based on the data in the table, it is clear that the proposed antenna is the most compact in size and has the best ECC level. It also seems that the isolation level between antenna ports is one of the best. Although the designs given in [25, 26] have higher peak gain, the proposed dual-port antenna is also quite good in terms of peak gain compared to the others. Finally, the proposed antenna system has a simplier and less costly design configuration.

Table 2 Comparison of the dual-port PMLPMA antenna with related 28 GHz antennas
Full size table

6 Conclusion

This article presents a novel log-periodic dual-port antenna system for 28 GHz 5G communication systems. The proposed dual-port antenna configuration consists of two identical PMLPMAs and each PMLPMA consists of 5 innovative mid-curved radiating log-periodic elements. In addition, a new octagonal ring-shaped FSS layer is used on the dual-port PMLPMA antenna to achieve high gain, and a detailed analysis of the FSS is also presented. The proposed antenna system demonstrates a simulated bandwidth ranging from 26.5 to 29.5 GHz. Measurements were also performed in a semi-anechoic chamber to validate the antenna performance, and the results closely align with the simulations, indicating the accuracy of the proposed design. The isolation levels between the ports were found to be less than − 30 dB in both simulated and measured results, ensuring good isolation and reduced interference. Moreover, the measured peak gain of the proposed antenna exhibited a notable improvement from 3 to 6.8 dBi, while the simulated peak gain increased from 3.5 to 7.35 dBi. Remarkably, the gain was enhanced by over 65% across the entire operating band, signifying the significant gain improvement achieved through the proposed design. These results validate the effectiveness of the novel log-periodic dual-port antenna, making it a promising candidate for 5G communication systems operating at 28 GHz frequencies.