1 Introduction

In the present scenario, no one can imagine a life without mobile phone which has touched our lives up to deep rooted level and with time there is an increasing demand for infinite bandwidth along with enormous data rate. So, researchers and innovators are constantly looking forward for a technique that can cope up with the hunger of this everlasting demand of channel capacity. Now multiple input multiple output (MIMO) technology is one of the major advancements that can give a hope in this challenging situation.

Wallace (Wallace et al. 2003) suggested that multiple input multiple output (MIMO) techniques can be used to enhance the channel capacity with no requirement of additional power because of its ability to transmit data over multiple channels. Furthermore, using MIMO systems pattern and polarization diversity can be achieved along with making the multipath fading less severe (Kaiser et al. 2009; Wang et al. 2009). Designing of MIMO antennas really takes a toll on the parts of researchers as there are several issues which are to be taken care of while designing. For an efficient MIMO antenna system, there are few performance matrices which are required in addition to regular single antenna systems such as envelope correlation coefficient (ECC), diversity gain (DG), total active reflection coefficient (TARC) and mean effective gain (MEG). Actually, effect of adjacent placement of antennas on overall system performance is the key factor that decides the above said parameters. However, UWB MIMO systems suffers from a major drawback of strong mutual coupling between adjacent antenna elements which severely affects the performance of the system. Hence main demand in MIMO antenna design is of enhanced isolation between adjacent antenna elements so for good performance of MIMO antenna systems, high isolation and low correlation coefficient is must (Mak et al. 2009; Saxena et al. 2018).Various methods have been suggested by researchers to reduce mutual coupling such as floating parasitic decoupling structure (Zhang et al. 2009) and a tree like structure (Srivastava and Mohan 2015) to enhance wideband isolation.

Cross-polarization is also a salient feature that should be taken into account while designing a single or MIMO system antennas. It is the polarization orthogonal to the normal polarization the antenna is meant to radiate. It should as low as possible. Over the past several years, researchers have designed multiple techniques for mutual coupling reduction and cross-polarization suppression. One such technique is using Defected Ground Structure (DGS), which involves the creation of a defect in ground plane intentionally. Specific geometrical pattern (single/periodic/aperiodic) is created in the ground plane of printed circuits. DGS was directly integrated with microstrip patch antenna and presented for the radiation pattern and other improvements (Khan et al. 2014). This perturbation created causes a non-uniformity and discontinuity in ground plane. A disturbance is created by DGS in the ground plane current and causes a variation in capacitive and inductive characteristics of transmission line such that multiple parameters are affected by introduction of DGS in ground plane. In recent years, the Improvement in cross-polarization characteristics of microstrip antenna using DGS has been proposed (Khandelwal et al. 2017; Khandelwal et al. 2014). Four-port MIMO antennas embedded with DGS also have been presented with diversity characteristics (Sharma et al. 2018; Kumar et al. 2019).

This work discusses a novel design of dual tapered patch supported by a similar defected ground structure which gives a wide bandwidth and a high stable gain. Finally, in next design innovation, a four element MIMO system is designed and fabricated and a parallel slits DGS pattern is introduced which gives high isolation, low correlation coefficient and suppressed cross-polarization with high gain.

2 Antenna structure and analysis

Figure 1 represents the schematics of top and bottom view of proposed antenna. Single element antenna is shown in Fig. 1a which is referred as Ant_1 in this study. It consists of a particular Y shaped patch, loaded with two side triangular structures fed by a microstrip line of 50 Ω characteristic impedance. The ground plane of the above structure is defected with a similar central shape as that of patch, shown in same figure. Ant_1 is offering an impedance bandwidth of 6.2 GHz ranging from 22.8 to 29 GHz. Ant_1 is also analyzed with and without DGS and triangular parasitic patches, respectively. Proposed structure without DGS and without both triangular parasitic patches is resonating at 23.9 GHz whereas it starts to offer two frequency bands at 23.9 GHz as wells as 27.5 GHz after embedding the DGS in the same structure (with DGS and without triangular parasitic patches). Thus, it is concluded that; (1) main central structure is resonating near about 23.9 GHz, (2) with the help of DGS it also starts to resonate at 27.5 GHz including previous band, (3) triangular parasitic patches maintains the return loss level below the value of − 10 dB between the frequency range of both above specified bands. S11 characteristic of Ant_1 and its evolutionary steps are depicted in Fig. 2.

Fig. 1
figure 1

Schematic of top and bottom view of proposed antenna; aAnt_1, bAnt_2, cAnt_3

Full size image
Fig. 2
figure 2

Simulated and measured S11 variation with frequency of Ant_1 and its evolutionary steps

Full size image

Further, four-element structure is designed as a MIMO antenna on a common ground plane and structure is referred as Ant_2. Schematic of top and bottom view of Ant_2 is displayed in Fig. 1b. Ant_2 consists of four elements (both the patch as well as DGS of each element are same as Ant_1) with broadside placement of adjacent antennas with a distance d. Ant_2 offers almost similar bandwidth of about 6 GHz ranging from 23 GHz to 29 GHz as shown in Fig. 3. It offers maximum mutual coupling from port#1 to port#2, port#3 and port#4 below the value of − 25 dB, − 35 dB, and − 38 dB, respectively.

Fig. 3
figure 3

Simulated and measured S-Parameter variation with frequency of Ant_2

Full size image

Moreover, three slits of width Wt, length Lt1, Lt2 and Lt1 are embedded in the ground plane of Ant_2 at a distance Wd, respectively. After embedding the slits in the ground plane, new structure is referred as Ant_3. Schematic of top and bottom view of Ant_3 is shown in Fig. 1c and the values of all the variables shown in schematic are listed in Table 1. Mutual coupling between each port is suppressed about 4 dB by embedding the slits in the ground plane while S11 remains almost same. Simulated and measured S-parameter variations of Ant_3 is shown in Fig. 4. Ant_3 offers almost same frequency band with maximum mutual coupling of − 28 dB, − 36 dB, and − 42 dB from port#1 to port#2, port#3, and port#4, respectively. Few S-parameters of Ant_2 and Ant_3 are observed similar to each other thus not shown in Figs. 3 and 4 to reduce the complexity of the graph. Such similar parameters are presented in following groups (1) S11, S22, S33, S44, (2) S12, S21, S23, S32, S34, S43, (3) S13, S31, S24, S42, S13, S31, and (4) S14, S41.

Table 1 Parameter specification of proposed antenna
Full size table
Fig. 4
figure 4

Simulated and measured S-parameter variations with frequency of Ant_3

Full size image

3 Result and discussion

RT-Duriod-5880 substrate of dimensions L and W with thickness h = 0.762 mm and relative permittivity εr = 2.2 is used for the fabrication of the proposed antennas. Fabrication is done by photolithography and measurement is done by Agilent™ PNA-L series Network Analyser and Agilent™ Spectrum Analyzer. The experimental characteristics of proposed antennas are measured in Anechoic Chamber. Fabricated prototypes of proposed antennas are displayed in Fig. 5. SMA connector having ground plane of about 14 mm width and 5 mm height is used to excite the Ant_1, this ground plane of SMA connector is worked as a metallic reflector placed in YZ plane which reflect back the backward radiations in forward direction and enhance the value of front to back ratio. Microstrip antennas with reflector also have been proposed previously in Sharma et al. (2018) and Kumar et al. (2019). On the other hand, the width of ground plane of SMA connector is reduced to the value of 5 mm with same height of 5 mm to maintain the minimum separation distance d between adjacent elements. The effect of SMA connectors are also considered in simulation of proposed structures.

Fig. 5
figure 5

Prototypes of fabricated antennas, a top view of Ant_1, b bottom view of Ant_1, c top view of Ant_2 (similar to Ant_3), d bottom view of Ant_2, e bottom view of Ant_3

Full size image

Figure 6 shows the gain variation with frequency of proposed antennas in which maximum gain of about 7.3 dBi, 7.1 dBi and 7 dBi is achieved at 28.2, 28.5, and 28.6 GHz of Ant_1, Ant_2, and Ant_3, respectively. After implementing MIMO elements in broadside direction, maximum gain is reduced slightly however variation between minimum and maximum values are reduced. In the operating frequency range, the gain varies from 5.1 to 7.3, 5.8 to 7.1, and 5.9 to 7 dBi of Ant_1, Ant_2, and Ant_3, respectively.

Fig. 6
figure 6

Simulated and measured gain variation of proposed antenna

Full size image

The normalized radiation pattern of Ant_1, Ant_2, and Ant_3 are shown in Figs. 7, 8, and 9, respectively. For comparison purpose co-polar and cross-polar radiation pattern of all three antenna structures are taken at common working frequency 27.5 GHz in both major E and H planes. Cross-polarization level of all the three antennas are found below the value of − 20 dB which is further suppressed by integrating DGS slits in Ant_3.

Fig. 7
figure 7

Simulated and measured radiation pattern of Ant_1 at 27.5 GHz of; aE-Plane, bH-Plane

Full size image
Fig. 8
figure 8

Simulated and measured radiation pattern of Ant_2 at 27.5 GHz of; aE-Plane, bH-Plane

Full size image
Fig. 9
figure 9

Simulated and measured radiation pattern of Ant_3 at 27.5 GHz of; aE-Plane, bH-Plane

Full size image

For MIMO structure, there are few parameters of interest such as envelope correlation coefficient (ECC), diversity gain (DG), total active reflection coefficient (TARC) and mean effective gain (MEG). Envelope correlation coefficient (ECC) is a measure of the amount of correlation between adjacent antenna elements and should be zero ideally while < 0.5 practically. For the proposed MIMO antenna ECC is measured by Agilent™ PNA-L series Network Analyser and is calculated on the s-parameters without including the efficiency.

$$ ECC = \frac{{\left| {S_{ii}^{*} S_{ij} + S_{ji}^{*} S_{jj} } \right|^{2} }}{{\left( {1 - \left| {S_{ii} } \right|^{2} - \left| {S_{ji} } \right|^{2} } \right)\left( {1 - \left| {S_{jj} } \right|^{2} - \left| {S_{ij} } \right|^{2} } \right)}} $$
(1)

Here i and j represent the respective port of the MIMO antenna.

DG describes the increase in signal to interference ratio on applying some diversity scheme. Mathematically it can be calculated using Eq. (2)

$$ DG = 10\sqrt {1 - ECC^{2} } $$
(2)

As low value of ECC is a measure of high isolation between adjacent ports and is measured as less than 0.01 while DG greater than 9.95 dB for port 1 with respect to ports 2, 3 and 4 respectively as shown in Fig. 10. Similarly, for port 2 with respect to ports 3 and 4 as well as for port 3 with respect to port 4 are shown in Figs. 11 and 12, respectively.

Fig. 10
figure 10

Measured ECC and DG for port 1 of Ant_3

Full size image
Fig. 11
figure 11

Measured ECC and DG for port 2of Ant_3

Full size image
Fig. 12
figure 12

Measured ECC and DG for port 3 of Ant_3

Full size image

MEG is a measure of relative mean power levels between signals transmitted from each antenna. Actually, it gives the average power received by test antenna to the sum of average powers that would have been received using two isotropic radiators, vertically and horizontally polarized respectively. The MEG can be calculated using Eq. (3)

$$ MEG = \mathop {\iint }\limits_{00}^{2\pi \pi } \{ \frac{XPR}{1 + XPR}G_{\theta } \left( {\theta ,\emptyset } \right)P_{\theta } \left( {\theta ,\emptyset } \right) + \frac{1}{1 + XPR}G_{\emptyset } \left( {\theta ,\emptyset } \right)P_{\emptyset } \left( {\theta ,\emptyset } \right)\} \sin \theta d\theta d\emptyset $$
(3)

Here XPR represents the cross-polarization power ratio and Gθ, Gϕ, Pθ and Pϕ are the elevation and azimuth components of antenna power gain patterns and angular diversity of incident power respectively. The practical value of ratio |MEGi/MEGj| should be < ± 3 dB, where i and j are antenna elements, and the ratio of MEGs of two antenna elements verses frequency lies in permissible range as shown in Fig. 13.

Fig. 13
figure 13

Ratio of MEGs of every two-antenna element for the proposed MIMO antenna

Full size image

TARC is given by square root of the ratio of the difference of total available power and the radiated power to total available power as given by Eq. (4)

$$ TARC = \sqrt {\frac{available power - radiated power}{available power}} $$
(4)

The practical value of TARC always lies between 0 and 1 which means less than 0 dB. It is mainly used to observe the effective return loss of the MIMO system. The mathematical expression of TARC is given by Eq. (5)

$$ TARC = \sqrt {\frac{{\left( {S_{ii} + S_{ij} } \right)^{2} + \left( {S_{ji} + S_{jj} } \right)^{2} }}{2}} $$
(5)

The variation of TARC verses frequency for proposed MIMO antenna Ant_3 is represented by Fig. 14 and it is observed that it is mostly less than − 5 dB.

Fig. 14
figure 14

TARC verses frequency for the proposed MIMO antenna

Full size image

A comparison analysis of proposed antenna Ant_3 with similar literature is presented in Table 2 and it is analyzed that proposed antenna offers better characteristics in terms of compactness, minimum mutual coupling, higher gain and low cross-polarization level with best diversity conditions among all the four ports.

Table 2 Comparison of proposed and earlier reported antennas
Full size table

4 Conclusion

Four-element MIMO antenna is successfully designed and analyzed. Proposed antenna is operating from 23 to 29 GHz and offers better radiation characteristics with suppressed cross-polarization below the value of − 20 dB within the whole operating band. A maximum gain of 7.1 dBi is noticed at 28.8 GHz whereas mutual coupling is achieved below the value of − 28 dB. Further, diversity characteristics are verified in terms of ECC, DG, TARC and ratio of MEG which lie within the specified limit for a MIMO antenna. All the simulated results are verified by experimental measured results and both results are found in good agreement.