Abstract
A compact, dual-port, triband MIMO antenna is designed and tested for three sub-6 GHz WLAN bands for IoT applications. The size and performance of the antenna make it versatile for the emerging IoT applications communicating using the 2.4 GHz, 5.2 GHz, and 5.8 GHz WLAN frequencies. The single radiating element of antenna geometry includes a pair of modified and optimized rectangular patches. A complimentary split ring resonator (CSRR) structure is incorporated in the ground plane as a defect (defected ground structure) to attain the third operational band. The orientation of the elements, orthogonally arranged, along with DGS, enhances the isolation between the radiating elements keeping it below 18 dB throughout the three operating frequency bands. The size of the final antenna is as small as mm2 build on commercially available and cheap FR4 substrate of thickness 1.5748 mm. A peak gain of 7.17 dBi is attained for the proposed MIMO system. This MIMO antenna also satisfies the diversity parameter requirements including , , dB, and bits/s/Hz. This makes the proposed antenna a good candidate for IoT applications employing WLAN frequencies.
1. Introduction
The present decade has seen an overwhelming interest in IoT technology, which is an extension of worldwide web. The web now includes all devices connected to the internet through the “Internet of Things” technology. The efficient and reliable communication between all the devices connected to the internet has ensured a massive number of applications where its benefits can be exploited. The applications include road traffic control and vehicular IoT [1–3], industrial IoT [4–6], and medical IoT [7, 8]. This further pushed the antenna researchers to provide wireless solutions which can provide a functional platform for more than one application [9]. Some major requirements of IoT applications are to maintain a good quality of signal, device compatibility, multifunctionality, and reliability.
WLAN is the wireless solution which enables flexibility and user mobility [10]. WLAN is also one among the most common communication technology which is profoundly applied in IoT applications. The operational bands for WLAN include 2.4 GHz, (2.4-2.484 GHz), 5.2 GHz, (5.15-5.25 GHz), and 5.8 GHz, (5.725-5.825 GHz). The lower band covers the low-bandwidth applications within medium range and the upper bands supports low-range, higher-bandwidth applications [11]. Compared to the traditional communication solutions, MIMO systems are the future of present-day communication technology despite its additional cost and complexity. The three significant features of MIMO, which makes it unique and popular are its beam steering capability, the enhanced data capacity, and the diversity characteristics [12, 13]. The major challenge faced in the modern-day MIMO design are in integrating multiple radiating elements in limited space with good isolation and attaining a low correlation between the closely placed antenna elements. Inclusion of lower WLAN band greater influences the size of the antenna, and additionally, a MIMO system needs to have enough spacing between the radiation elements to avoid mutual coupling between the elements [14–17].
Inclusion of the lower WLAN frequency while maintaining the compactness and isolation in a MIMO antenna system is a hard task for an antenna design engineer. The antennas in the literature includes the lower band at the cost of size [18–22] or low isolation [23]. [24] refers to a miniaturized antenna of size for upper range of frequency (mm range of 35 GHz), and [25] proposes the use of decoupling structure ( dB) in the design of a closely placed MIMO antenna elements for 5.4 GHz of size . The novelty of the proposed structure is the achievement of a compact size and high isolation (>18 dB for lower band and >23 dB for upper bands), irrespective of inclusion of the lower band of operation (2.4 GHz). This paper proposes a compact 2-port, triband MIMO antenna for WLAN IoT applications with high isolation, simple and compact structure, good gain, and radiation characteristics. The fine tuning of the antenna to 2.4 GHz, 5.2 GHz, and 5.8 GHz is an additional advantage of this design.
2. Antenna Design and Dimensions
The evolution stages of the design in shown in Figure 1, and the single radiating element of the 2-port MIMO system is as shown in Figure 2. Low-cost easily available FR4 substrate is used for the fabrication, and high-frequency structure simulator (HFSS) software is used for the design and analysis of the antenna. The overall dimension of the final antenna is .


It consists of five stages as shown in Figure 1. The study starts with stage-1, in which, a rectangular microstrip antenna of size is designed at WLAN frequency.
The dimensions of rectangular patch antenna are calculated from the following equations: where is the width of the patch, is the speed of light in free space, is the design frequency, and is the substrate’s dielectric constant.
The length of the rectangular base structure is calculated as
where is the length of the patch, is the wavelength in free space, and is the fringing length. The effective dielectric constant of structure is calculated by (3), and Fringing length is calculated as in (4).
The width of the feedline is calculated from Equation (5) for 50 characteristics impedance () as where
The basic patch gave no significant bandwidth over the minimum threshold of -10 dB return loss. The second step added another rectangular patch to the right side of the first patch. This helped in achieving a single narrow band (1.922 GHz–2.009 GHz) resonating at 1.953 GHz with a return loss of 23.71 dB. An additional band is observed around 5 GHz but it is not below -10 dB. In the following step, an addition connect was offered to the first and second rectangular patches. This helped in shifting the bandwidth to the right to (2.18 GHz-2.286 GHz) resonating at 2.235 GHz with -36.244 dB. A shift is observed in the additional bandwidth seen in step two but the return loss could not be enhanced in this step. In step 4, a square slot of was cut in the right corner of the first rectangular patch. This provided an enhanced return loss to the additional operation band. The final resonating element has dual operational bandwidths, 2.207 GHz-2.28 GHz and 4.849 GHz-5.249 GHz resonating at 2.235 GHz and 5.04 GHz, respectively. The summary of for the evolutionary stages is given in Figure 3.

The parametric study is conducted on every element added in each step of progress to the final radiating element. In step 2, the length and position of the second rectangular patch added is optimized through parametric study as shown on Figure 4(a). The results of the study on step two is given in Figures 4(b) and 4(c).

(a)

(b)

(c)
In step 3 of the design of radiating element, a rectangular stub is used to make an extra connection with the first and second rectangular patch. The width of this patch is set under parametric study to conclude that the best return loss characteristics is attained at 2.4 GHz when the width of the stub is 1.5 mm. The diagrammatic representation of the parametric analysis done on step 3 is shown in Figure 5.

In the final step of the design, a slot is embedded in the first rectangular patch. Embedding this slot results in two operating bands. The first band ranged between 2.207 GHz and 2.28 GHz, and the second band from 4.84 GHz to 5.24 GHz. The length and width of the slot embedded are again optimized through parametric study. Figure 6(a) shows the parameters investigated under this study, Figure 6(b) shows the parametric analysis on length of the slot, and Figure 6(c) shows the parametric study on the width of the slot.

(a)

(b)

(c)
The radiating element was thus designed. Further, a 2-element MIMO system was designed as shown in Figure 7. Individual radiating elements were organized in an FR4 substrate of size . The MIMO antenna with triband evolved through three stages, stage 1 is named Ant_I and is shown in Figure 7(a). In the next stage, two-element MIMO is configured by placing the elements orthogonally to use antenna utilization to enhance isolation. This antenna is names as Ant_II and is shown in Figure 7(b). In stage 3, a CSRR is embedded into the ground plane to achieve the third band of resonance and improve isolation. Stage 3, with CSRR in DGS, is named as Ant_III as shown in Figure 7(c). The final top and bottom view of the antenna is shown in Figures 7(d) and 7(e), respectively. The optimized design dimensions are summarized in Figure 8(a) and Table 1.


(a)

(b)
CSRR is used as an effective method to achieve compactness while designing for a particular frequency. CSRR can be excited by means of an axial time varying electric field or by means of a magnetic field applied to the plane of antenna. Under ideal conditions, i.e., a perfectly conducting and zero-thickness metallic surface, it can be demonstrated that the resonance frequency of the CSRR is identical to the resonance frequency of the SRR (provided identical dimensions and substrate are considered). Indeed, the capacitance of the CSRR, , is the capacitance of a disk of radius surrounded by a ground plane at a distance of its edge. The inductance is given by the parallel connection of the two inductances of the metallic strips connecting the inner and outer metallic regions of the CSRR. These inductive values are given by , where , with being the per unit length inductance. Due to change in equivalent inductance and capacitance of the ground plane, effective current distribution of ground plane changes resulting in the appearance of a new ISM band (5.119 GHz–5.353 GHz). The CSRR parameters include the average ring of radius , i.e., the mean radius of the ring, and ring width as shown in Figure 8(b).
CSRR characteristics have been widely exploited by antenna designers to achieve various desired parameters. It has seen to be used in literature to achieve lower resonant frequencies in miniaturized antenna [26], to realize bandpass and bandstop filters [22, 27], to enhance isolation in MIMO antenna systems, [28], and so on. The circuit models corresponding to CSRR are not very accurate because circuit modal of unit cell must take into account the coupling between the resonator and the line, i.e., interresonator coupling. In addition to this, the decrease in electrical size of the CSRR is seen to have an effect in narrowing the operational bandwidth.
Figures 9 and 10 picturize the operational bandwidths and isolation achieved in each stage of evolution. Ant_I shows two operational bandwidths, 2.207 GHz to 2.28 GHz, resonating at 2.235 GHz (-18.48 dB) and 4.84 GHz to 5.04 GHz, resonating at 5.04 GHz (-22.31 dB). Ant_II is the orthogonally placed 2-element MIMO system with two bands of 2.299 GHz to 2.360 GHz and 5.806 GHz to 6.334 GHz. The return loss of the second band originally present in the simulation results of Ant_I was reduced below the reference level of -10 dB. So, the two bands of Ant_II resonated at 2.32 GHz (-11.45 dB) and 6.121 GHz (-18.879 dB), respectively. The addition of CSRR as a DGS to Ant_II attained enough return loss (>-10 dB) for the second band of Ant_I. Ant_III, therefore, has three bands of operation as 2.336 GHz to 2.449 GHz, resonating at 2.4 GHz (-20.17 dB); 5.696 GHz to 6.29 GHz, resonating at 5.8 GHz (-40.03) similar to that of Ant_II; and the third additional band (as in Ant_I) 5.119 GHz to 5.353 GHz, resonating at 5.2 GHz (-33.16 dB).


Figure 10 shows the isolation between the elements in Ant_II (without CSRR) and Ant_III (with CSRR). For band 1 (2.336 GHz–2.449 GHz), the isolation ranges between -15.97 dB and -18.87 dB in Ant_II and -24.05 dB and -25.94 dB in Ant_III. At the first resonance (2.39 GHz), the isolation in Ant_II is -16.32 dB and in Ant_III is -24.05 dB. For the second band (5.119 GHz–5.353 GHz), the isolation ranges between -32.94 dB and -36.54 dB in Ant_II and -41.36 dB and -25.38 dB in Ant_III. At the second resonance (5.235 GHz), the isolation in Ant_II is -35.32 dB and in Ant_III is -28.06 dB. For band 3 (5.696 GHz–6.29 GHz), the isolation ranges between -16.84 dB and -27.34 dB in Ant_II and -32.32 dB and -41.28 dB in Ant_III. At the third resonance (5.84 GHz), the isolation in Ant_II is -24.7 dB and in Ant_III is -37 dB. Figure 11 shows the image of the fabricated antenna.

(a)

(b)
3. Results and Discussions
A dual-element, triband MIMO antenna is designed and fabricated for measurements. The antenna offers three versatile frequencies of operation in the sub-6 GHz band for various WLAN and IoT applications. The return loss for the antenna was measured by a vector network analyzer (VNA), and the results seems to satisfactorily match with the simulated results. The performance of the MIMO antenna was also studied under an anechoic chamber to determine the gain and the radiation characteristics. Slight deviations are observed between the simulated and measured values which might be a possibility due to minute error that occurred during the fabrication, soldering process, or testing set up.
3.1. Parametric Study
A set of parametric study was conducted to attain best possible isolation between the radiating elements and to achieve all three bands of operation.
3.1.1. Orientation of the Antenna Elements
The element orientation in the substrate was carefully optimized to achieve maximum isolation. Figure 12 shows the different orientations investigated. The antenna elements were placed at 90° with respect to one another (orientation 1), at 180° with each other (orientation 2), and at 180° and moved to the corners to keep them maximum apart from each other (orientation 3). The orientation 1 gave good isolation at the lower band (2.33 GHz-2.44 GHz) but isolation was above -10 dB for the second band (5.11 GHz-5.35 GHz) and between -15 dB and -20 dB for the third band (5.69 GHz-6.29 GHz). Orientation 2, meanwhile, provided good isolation to (5.69 GHz-6.29 GHz) band. Orientation 3 attained lowest isolation for all the three resonant frequencies as shown in Figure 13.


The orientation of the CSRR in the ground plane of the MIMO antenna also influences the performance of the MIMO system as shown in Figure 14. Orientation I provides three operating bands but the return loss is very less for band (5.11 GHz-5.35 GHz and 5.69 GHz-6.29 GHz). Orientation II gives a slightly shifted version of the bandwidths towards the right with an additional resonance at 1.8 GHz. The resonant frequencies in orientation II are 1.88 GHz, 2.55 GHz, 5.39 GHz, and 6.31 GHz. The third orientation attains a band at 1.8 GHz but not enough to have a return loss beyond -10 dB. The first lower band resonates at 2.43 GHz, the second at 5.28 GHz, and the third band at 6.1 GHz. The orientation of CSRR as in orientation IV optimizes the resonant frequencies to 2.38 GHz, 5.23 GHz, and 5.84 GHz. Figure 15 shows the variations in according to the orientation of CSRR in the ground plane. At some points, it is seen that the isolation for orientation 4 is less compared to other orientation by orientation 4 shows better isolation than all other orientations for the operating bandwidths.

(a)

(b)

(c)

(d)

(e)

The other parameters of CSRR investigated parametrically are shown in Figure 16(a). The includes the width of the rectangular slot cut on the CSRR (Figure 16(b)), the width of the rings of the CSRR (Figure 16(c)), and the distance or spacing between the rings of the CSRR (Figure 16(d)). After the parametric analysis, the slot width of CSRR is fixed at 2 mm, and the width of CSRR and spacing between the rings of CSRR are fixed at 1 mm.

(a)

(b)

(c)

(d)
3.2. Antenna Parameters
The simulated and measured values of the antenna parameters for the proposed antenna is discussed below.
3.2.1. Parameters
The simulated and measured parameters are in good agreement with each other. Figure 17(a) shows the summary of both return loss and isolation values for the proposed antenna.

(a)

(b)
3.2.2. Gain, Efficiency, and Radiation Pattern
The gain and radiation pattern of the proposed MIMO antenna is measured in an anechoic chamber with the help of a particular antenna measurement setup. Figure 17(b) shows the plot of simulated and measured gain and radiation efficiency. Both these parameters explain how well the designed antenna performs. The maximum gain attained by the antenna is 7.17 dBi at 5.28 GHz. It is observed that the gain is lower at lower frequencies. This is because the antenna becomes smaller with respect to the wavelength at lower frequencies. Gain at 2.4 GHz is 1.893 dBi and at 5.84 GHz, gain attained is 6.71 dBi. The radiation efficiencies are 65.74% (2.4 GHz), 84.75% at 5.28 GHz, and 92.3% at 5.84 GHz.
The radiation pattern for the -plane and -plane at all the three resonant frequencies are as shown in Figure 18. Figure 18(a) is the radiation pattern at 2.4 GHz, Figure 18(b) is the pattern for 5.2 GHz, and Figure 18(c) is the radiation pattern at 5.8 GHz.

(a)

(b)

(c)
3.3. MIMO Diversity Parameters
MIMO antennas are analyzed with a set of diversity parameters as the envelope correlation coefficient (ECC), the diversity gain (DG), total active reflection coefficient (TARC), and channel capacity loss (CCL). Each of these has be evaluated below for the proposed dual-element, triband MIMO antenna.
3.3.1. Envelope Correlation Coefficient
When multiple antennas are used simultaneously, the correlation between the received signals helps in analyzing the performance of the system. The far field radiation characteristics helps in evaluating ECC by
The far field radiation characteristics of port 1 and port 2 are represented as and F2. Ideally, the preferred value of ECC is 0, but for practical MIMO antenna systems, ECC less than 0.5 is accepted as acceptable performance.
3.3.2. Diversity Gain
The diversity gain is another parameter of interest in MIMO systems. DG is easily calculated from ECC using Equation (8). The value of DG needs to be very close to 10 dB for the operating bands in a well performing MIMO.
The ECC and DG plot for the proposed MIMO is shown in Figure 19. Both the measured and simulated ECC are well below 0.5 for this MIMO antenna, and the DG for the operating bands are close to 10. ECC has also been calculated from far field radiation pattern as shown in Figure 19(b). Table 2 summarizes the ECC and DG values for all the three operational bands.

(a)

(b)

(c)
3.3.3. Channel Capacity Loss
CCL is the parameter used to measure the information transmission rate of the wireless channel and CCL should be less than 0.5 bits/sec/Hz for a good communication system. The channel capacity loss for a system is calculated from
where is the performance of the MIMO antenna and is given as in
The CCL values are 0.34 bits/sec/Hz for 2.39 GHz, 0.108 bits/sec/Hz for 5.23 GHz, and 0.261 bits/sec/Hz for 5.84 GHz. The plot of simulated and measured CCL is shown in Figure 20(a).

(a)

(b)
3.3.4. Total Active Reflection Coefficient
TARC is another major parameter to study the performance of a MIMO antenna system. TARC studies the diversity performance while multiple antennas are working simultaneously. TARC, in other words, can be considered the return loss of a MIMO antenna system. TARC can be calculated from the formulation given as
where is the difference in phase between the excitation signals.
For a well performing MIMO antenna, TARC is expected to be less than 0 dB. In this proposed antenna, TRAC is less than -18 dB for all the operational bandwidth which concludes that the proposed MIMO has good diversity performance. The simulated and measured TARC is shown in Figure 20(b).
Table 3 summarizes the current state of arts of similar multiband MIMO antennas.
3.4. Radiation Pattern Analysis
The Friis transmission equation relates the power received to the power transmitted between a transmitting and receiving antennas separated by a distance where is the largest dimension of either antenna. The Friis transmission formula is given as follows: where is the wavelength of free space in meter, is the receiving antenna power in dBm, is the transmitting antenna power in dBm, and and is the gain of the transmitting antenna and receiving antenna.
The representation of the set up to evaluate the performance of the antenna is shown in Figure 21. The experimental setup consists of the proposed transmitting microstrip antenna and a standard Yagi-Uda antenna as shown in Figure 22.


(a)

(b)

(c)
The above Friis equation can be represented as
For a given transmitted power () radiated by proposed antenna, the received power () is being measured and the plot of verses gives a straight line with slope equal to as per Equation (12). As the receiving antenna is standard Yagi-Uda antenna so its gain () is known whereas the gain of transmitting antenna is easily obtained from the gain plot of the proposed antenna.
The range can be obtained as per Equation (13) and confirms a wider coverage area as indicated by the radiation efficiency.
4. Conclusion
IoT technology has evaded almost every field in the present day, and wireless communication technologies plays a vital role in the effective implementation of IoT systems. Modules like ESP8266 are widely used for various IoT applications. Multifrequency operational antennas with enhanced performance and reduced size are desired for many IoT and modern wireless applications. The paper suggests a dual-element, triband MIMO antenna with a microstrip line feed. The antenna is low profile and compact and built of FR4 substrate of size . The simple and identical radiating patches with the CSRR embedded in the DGS provides three versatile operating frequencies of 2.4 GHz, 5.2 GHz, and 5.8 GHz. The antenna exhibits good isolation and stable gain and radiation characteristics and can be used for IEEE 802.11 a/b/g applications and other IoT systems connected to the above frequencies of operation. A peak gain of 3.56 dBi is attained for the proposed MIMO system. The MIMO diversity parameter requirements including , , dB, and bits/s/Hz. This makes the proposed antenna a good candidate for IoT applications employing WLAN frequencies. The antenna has also been tested with standard Yadi-Uda antenna to analyze its performance when connected to ESP8266 module.
Data Availability
The data will be available upon request.
Conflicts of Interest
The authors declare that they have no known competing financial interest or personal relationships that could have appeared to influence the work reported in this paper.