1 1. Introduction

The recent communication systems need a wide bandwidth, and high transmission data rate to improve the quality services of these modern systems. Therefore, the Federal Communications Commission (FCC) specified a frequency band from 3.1 to 10.6 GHz (7.5 GHz) to be utilized in UWB technology [1]. The antenna is an essential component in the UWB technology; so researchers made several attempts to obtain the best design to fulfill the UWB technology demands such as miniaturized size, low price, wide bandwidth, and omnidirectional patterns with constant gain [2,3,4,5,6,7].

Because of the multipath problem introduced by the UWB technology, the multi-input-multi-output (MIMO) technology is utilized to solve it. The technology of MIMO is used to enhance the capacity of the system; so the integration between the two technologies is the best choice for improving the performance of the system. The MIMO antenna should have a small size, high isolation between ports, low cost, and many elements that can be used. Achieving high isolation is the main challenge in the MIMO system. Researchers developed several attempts to enhance the isolation of the UWB MIMO antenna. Some of them used decoupling structures such as stubs [8,9,10] and the others arranged the antenna elements in an orthogonal layout [11,12,13,14,15].

There are two narrowband frequencies allocated for WiMAX technology around 3.6 GHz and WLAN technology around 5.5 GHz could be overlapped with the UWB bands. So, a band rejection mechanism such as resonators added at a specific location in the patch [16], ground [17], and close to the feed line [18] should be utilized to bypass this problem. UWB MIMO antenna with band rejection mechanisms is proposed in [12, 19,20,21,22]. In [12], 4 ports with band rejection mechanism using L- shaped stubs connected to the semielliptical patch radiator to avoid the band within 3–4 GHz (WiMAX) are introduced. A band-stop structure etched within the ground plane of the four elements UWB MIMO antenna to produce single band rejection is investigated in [19]. In [20], two L-shaped slits to introduce band rejection etched in the ground plane of two elements UWB MIMO are proposed. Quad elements with an L-shaped stub to generate band rejection is discussed in [21]. A stub attached to the ground plane of the 4-port UWB MIMO is utilized to create band rejection as displayed in [22].

In this paper, a compact size 4-port UWB MIMO antenna operated at a frequency band from 3 to 12 GHz and with single-band rejection within 3.3 GHz up to 4.3 GHz (WIMAX application) has been suggested. The suggested antenna elements are arranged perpendicularly to save space and increase the isolation between the elements. The four antenna elements have an overall size of 47 × 47 mm2. The MIMO diversity performance results have ECC < 0.005, DG 9.98 dB, and CCL< 0.4 bit/s/Hz within the operating bands. The suggested antenna is simulated using the CST suite, fabricated, and tested to confirm the simulated results.

2 Proposed single antenna structure

The 2D configuration structure of the UWB single antenna is illustrated in Fig. 1(a). The fabricated photo of the proposed antenna is shown in Fig. 1(b). The antenna is composed of a rectangular radiator with steps and curved edges on the top of the substrate to enhance the operating frequency band. The antenna has a partial ground plane with a length (Lg) of 6 mm and a rectangular slot of 3.5 mm× 3 mm etched in it to achieve the desired UWB behaviour. A simple meandered slot with a total length of λg/4 as calculated from Eq. (1) is etched in the rectangular radiator to reduce the interference with other devices working with WiMAX technology.

$$L_{total} = \frac{\lambda g}{4} = \frac{{\text{c}}}{{4f_{notch} \surd \varepsilon eff}}$$
(1)
Fig. 1
figure 1

Proposed single antenna structure a 2 D structure b Fabricated prototype

The low price FR4 substrate with 1.6 mm thick and 4.4 dielectric constant is proposed and utilized in the fabrication process.

Figure 2 illustrates the S11 simulated and measured results and the simulated result without using a slot in the radiator. From the simulated results, the antenna without a slot is worked with S11≤ − 10 dB at frequency band from 3.5 up to 12 GHz. While the antenna with a slot is operated from 3 up to 12 GHz with S11≤ − 10 dB instead of the region from 3.3 to 4.1GHz (the band notched) with a center band of the notch around 3.7 GHz and S11≥ − 10dB. From the measured results, the antenna is worked at a frequency band with S11≤ − 10 dB from 3 up to 12 GHz except for the frequency of the band notched from 3.3 to 4.3 GHz with a center band of the notch around 3.7 GHz. The two results have the same trend with a small shift between them because of the fabrication process and human errors due to the soldering.

Fig. 2
figure 2

Simulated and measured S11 results of the single antenna structure

There are four steps are executed to transfer the proposed antenna from the initial design to the proposed structure as shown in Fig. 3. The first step is the simple rectangular patch with a partial ground plane (Lg= 6 mm). The antenna is operated with S11≤ − 10 dB from 3.7 up to 5 GHz as shown in Fig. 3 (the dotted red curve). The second step with the same length as the ground plane and the patch has a step and curved edges. The antenna is worked with two bands from 3.7 to 4.8 GHz and from 7 up to 11 GHz (the dashed blue curve). The third step is the same as the second step except the ground plane has a rectangular slot etched in it. The antenna is operated in UWB operation from 3.5 up to 12 GHz (the dashed green curve). Finally, a simple slot is added to step three to achieve the desired band-notched feature as shown in Fig. 3 (The solid black curve). As well, the ground length (Lg) affects the antenna impedance bandwidth as shown in Fig. 4. The antenna is operated in a narrow band when full ground with Lg= 22 mm is introduced as shown in Fig. 4. By decreasing the ground length, the impedance bandwidth of the antenna is affected as shown in Fig. 4. When Lg=11 mm also the narrow band behavior around 14 GHz is presented. When Lg=6 mm dual-band operation is introduced. Finally, by etching a simple rectangular slot in the ground, the UWB feature can be achieved as shown in Fig. 4.

Fig. 3
figure 3

The simulated S11 results and the 2 D configuration of the UWB single antenna structure evolution

Fig. 4
figure 4

Simulated S11 results of UWB single antenna structure at different ground lengths (Lg)

The current distribution simulated results at 3.7 GHz and 8 GHz are illustrated in Fig. 5. The current is collected around the slot in the radiator at the frequency of the notch as shown in Fig. 5(a) which indicates that the antenna can't radiate at this frequency band. Fig. 5(b) illustrates that the antenna can radiate simply at 8 GHz (frequency away from the notch).

Fig. 5
figure 5

Simulated current distribution results of UWB single antenna structure a @ 3.7 GHz b @9 GHz

3 Two-port antenna

The previous single antenna has two orientations to achieve the two ports MIMO operation as shown in Fig. 6. The first one is side by side as shown in Fig. 6(a) and the second is the perpendicular orientation as shown in Fig. 6(b). We don’t use any decoupling structure to simplify the antenna structure. The distance between the edges of the two elements is the same as 9 mm (0.09 λ0 at 3 GHz) as shown in the figure. The simulated S21 is illustrated in Fig. 7. It is noticed that the perpendicular orientation has S21 lower than − 25 dB within the entire frequency band while the side by side orientation has S21 lower than − 10 dB, especially at the lower frequency band which means it has high coupling between elements [12,13,14,15]. So, the perpendicular orientation is used to form the proposed 4 ports MIMO as discussed in the next section.

Fig. 6
figure 6

Two ports UWB antenna layout a Side by Side orientation b Orthogonal orientation

Fig. 7
figure 7

S21 results of the two ports UWB antenna configurations

4 Four-port proposed antenna

4.1 Layout and S-parameter results

The 2D layout structure of the proposed 4-port antenna is shown in Fig. 8. The antenna has a miniaturized size of 47mm×47 mm. The antenna elements are arranged based on the previous investigation. The antenna at port 2 has a perpendicular orientation to the antenna at port 1, port 3, and the opposite direction with the same orientation as the antenna at port 4. As well as, the antenna at port 3 has the same orientation in opposite direction as the antenna at port 1 and perpendicular orientation with the antenna at port 2, port4. The antenna elements' separation between edges equals 10 mm (0.1 λ0 at 3 GHz) as shown in Fig. 8(a). The structure has a symmetry property so; the results from port 1 are extracted only. The MIMO antenna is fabricated and its fabricated prototype photo is illustrated in Fig. 8(b).

Fig. 8
figure 8

Layout of the proposed four-port UWB antenna a 2 D structure b Fabricated prototype

The four ports VNA (R&S ZVB 20) is used to test the proposed antenna by gauging the S-parameters to validate the antenna performance. Figure 9 displays the S11 simulated and gauged results. It is noticed that the measured result has S11 ≤− 10 at frequency band (UWB) from 3 to 12 GHz except the band notched band from 3.3 to 4.3 GHz. Figure 10 shows the simulated and gauged results of S21, S31, and S41 (at port 1). The measured results show that the coupling between antenna elements is lower than − 22 dB between antennas (1, 2), antennas (1, 4) (perpendicular orientation), and lower than − 18 dB between antennas (1, 3) (same and opposite orientation). This means the proposed antenna has high isolation without using isolation structures. Finally, the gauged and simulated results matched well with small deviation because of fabrication tolerance and process.

Fig. 9
figure 9

S11 results of the proposed four-port UWB antenna @port 1

Fig. 10
figure 10

S21 results of the proposed four-port UWB antenna @port 1

Figure 11 shows the simulated current on the surface of the proposed antenna at 3.7 GHz and 9 GHz. As illustrated in Fig. 11(a) the current has a high-level concentration around the slot which enables the antenna to block the radiation at this band. Figure 11(b) displays the distribution of current at 9 GHz. The current has a high-level concentration around the radiator which enables the antenna to radiate at this frequency also there is a small level of current is transmitted to the other three ports which confirms the high isolation between them.

Fig. 11
figure 11

Current distributions of the proposed four-port UWB antenna @port 1 a @ 3.7 GHz b @9 GHz

4.2 Radation patterns

The proposed antenna radiation patterns (at port 1) are measured inside an anechoic chamber and the measuring setup is shown in Fig. 13. The radiation patterns in two planes (x-z and y-z) are measured at 3.7 GHz, 7 GHz, and 9.5 GHz, and their normalized results are shown in Fig. 12 when the other three ports are connected with a matched load. As shown, the antenna has semi omnidirectional and semi bidirectional at the x-z plane and y-z plane respectively. Also, the level of the pattern at 3.7 GHz (notched frequency) is low. Finally, the simulated and the measured results have the same tendency to validate the antenna performance. Figure 13 displays the peak gain of the antenna. The antenna has gauged peak gain ranging from 2 to 4 dBi from 3 up to 12 GHz. But the antenna has a peak gain of around − 4 dBi at the frequency band of the notch as shown in Fig. 13.

Fig. 12
figure 12

Normalized radiation pattern results of the proposed four-port UWB antenna @port 1 a @ 3.7 GHz b @ 7 GHz c @ 9.5 GHz

Fig. 13
figure 13

Peak gain results of the proposed four-port UWB antenna @port 1

4.3 4.3 MIMO analysis evaluation

The main three parameters to validate the proposed MIMO performance are ECC, DG, and CCL. The ECC determines the connection between MIMO antenna elements. The ECC with a small value measures the high achievement of the MIMO system. The ECC can be extracted from the far-filed results as Eq. (2) [23, 24].

$$ECC = \rho_{e} = \frac{{\left| {\smallint \smallint 4\Pi \left[ {F_{1} \left( {\theta ,\varphi } \right) \cdot F_{2} \left( {\theta ,\varphi } \right)d\Omega } \right]} \right|^{2} }}{{\smallint \smallint 4\Pi \left| {F_{1} \left( {\theta ,\varphi } \right)} \right|^{2} d\Omega \smallint \smallint 4\Pi \left| {F_{2} \left( {\theta ,\varphi } \right)} \right|^{2} d\Omega }}$$
(2)

where, the achieved radiation patterns are Fi (θ, Ф).

The ECC simulated and measured results at port (1,2), port (1,3), and port(1,4) are shown in Fig. 14. The measured ECC results have a value lower than 0.005 at all frequency bands except at the notched band equals around 0.01. Also, good matching between the simulated and measured results has been accomplished. The value of measured ECC is lower than the accepted limit of ≤ 0.5.

Fig. 14
figure 14

ECC of the proposed four-port UWB antenna @port 1

Figure 15 illustrates the measured and simulated DG results which are extracted from Eq. (3) [25].

$${\text{DG}} = 10 \times \sqrt {1 - \left| {{\text{ECC}}} \right|}$$
(3)
Fig. 15
figure 15

DG results of the proposed four-port UWB antenna @port 1

The DG has a value higher than 9.98 dB at all frequency bands while equaling 9.96 at the notched band. Finally, the transmission data rates through the channel can be evaluated using CCL (bit/s/Hz). The value of the CCL can be extracted using Eqs. (4), and (5) [26].

$$C(Loss) = - \log_{2} \det (\psi^{R} )$$
(4)
$$\begin{gathered} \psi^{R} = \left[ {\begin{array}{*{20}c} {\rho 11} & {\rho 12} \\ {\rho 21} & {\rho 22} \\ \end{array} } \right],\rho_{ii} = 1 - \left( {\left| {S_{ii} } \right|^{2} + \left| {S_{ij} } \right|^{2} } \right) \hfill \\ {\text{and}} \hfill \\ \rho_{ij} = - \left( {S_{ii}^{*} S_{ij} + S_{ji}^{*} S_{ij} } \right),fori,j = 1or2 \hfill \\ \end{gathered}$$
(5)

Figure 16 displays the simulated and gauged results of CCL. The CCL has a value lower than 0.4 bit/s/Hz (the acceptable limits) within the entire frequency band excluding the notched band. In the end, we compare our design with others to validate our novelty. It is clear that our antenna has a compact size with good MIMO performance which promotes our design to be suitable for UWB MIMO systems (Table 1).

Fig. 16
figure 16

CCL results of the proposed four-port UWB antenna @port 1

Table 1 Comparison of the suggested design with other previous designs

5 Conclusion

A small size 4-port UWB MIMO antenna with single-band rejection has been suggested. The simple meandered slot etched in the radiator has been utilized to mitigate the interference with the surrounding WiMAX signal. The suggested MIMO has been placed on 47 × 47 × 1.6 mm3 FR4 substrates. The low mutual coupling between elements has been realized by forming the antennas orthogonally without adding decoupling elements. The antenna has been worked within 3 GHz up to 12 GHz with > 22 dB isolation and band rejection band within 3.3 GHz up to 4.3 GHz. The suggested antenna has omnidirectional radiation patterns with a peak gain of 3.5 dBi except for the band rejection bands the peak gain has − 4dBi. The MIMO diversity performance has been discussed and achieved good results with ECC < 0.005, DG 9.98 dB, and CCL< 0.4 bit/s/Hz within the operating bands. The suggested antenna is compared with other previous antennas, and the outcomes illustrate that the suggested antenna is a good choice for band rejection UWB MIMO applications.