Design and Study of a Miniaturized Millimeter Wave Array Antenna for Wireless Body Area Network

A miniaturized millimeter wave (mmWave) antenna for wireless body area networks is proposed in this paper. The antenna is found to be operational in the V-band, around the 60GHz frequency range, with high efficiency of up to 99.98% in free space simulations. A multilayer, thin substrate was implemented in the design to enhance radiation efficiency and gain. The antenna seems to be most suitable for small electronic devices and wireless body area network (WBAN) applications because of its low profile and lighter weight concept. To enhance its performance, several arrays of different orders were created. The Parallel-Fed and Tapered Feed Line methods were followed to design the planar arrays with 1 × 2, 1 × 4, and 2 × 2 elements in the primary design. Free space results were compared, and a 2 × 2 element array was found to be the most balanced according to the simulations. To justify the eligibility of these designs for WBAN applications, a virtual human body model was created within the 3D computer-simulated environment and the simulations were repeated, where four equal-spaced distances were taken into account to identify the antenna and its array behavior more accurately. Simulations returned optimistic results for the 2 × 2 element planar array arrangement in almost all parameters, even when placed close to the human body at any distance greater than 2mm.

priorities. eir compact size, lightweight configuration, and simplified structures make them most suitable for ultrawideband (UWB) or millimeter wave (mmWave) antennas. While achieving higher gain is also a challenge for small form factor (SFF) antennas like microstrip antennas, this issue can be solved by arranging antennas in arrays. ere are different techniques for placing antenna elements in an array. A common classification, which divides the arrays into two types based on how they are arranged, is the series-fed array, while the other is the parallel-fed array antenna. Series-fed antennas are those antennas whose elements are connected in series with a single transmission line fed from the excitation. On the other hand, a parallel-fed antenna has multiple radiating elements connected by separate feeding lines. Different techniques have their own specific benefits. For example, parallel-fed antennas have a wider bandwidth than series-fed arrays, with 10% of the operating frequency compared to 1-3% of the operating frequency of series-fed ones [7]. A combination of these two is also available [8][9][10]. Whatever the technique is, the ultimate goal for making antenna arrays is to achieve one or more better antenna parameters.
Apart from other positive factors, microstrip antennas offer narrow bandwidth and lower gains. However, researchers investigated coupling methods to overcome these drawback issues by implementing antenna arrays. One of the prominent applications of these microstrip antennas is the wireless body area network (WBAN), where essential human physical data is collected with the help of placing several nodes on the body in various locations. ese nodes gather this information and send it to the base station or the server with the help of a master node. WBAN or WBCN (Wireless Body-Centric Network) is mostly used for patient status monitoring in healthcare centers or monitoring an athlete's vital health signs in sports. To operate these tiny electronic devices wirelessly, embedded antennas must be lightweight, power-efficient, and small in size, making microstrip antennas the best candidate for the job. Manufacturing microstrip antennas is also budget-friendly, less complicated in structure, and much more convenient to integrate. e size of the antennas remains pretty small even when they are arranged in lower-order arrays, i.e., 2 × 2 or 4 × 1, etc. An antenna for a wireless body area network needs to be compact in size, printed, and flexible in design. e antenna also needs to have very good performance parameters to establish power-efficient communication links over the body-mounted wearable nodes. e antenna also needs to have a large bandwidth for high data rate communications. In a wireless body area network, the antenna also needs to have a high gain to allow communication with other colocated devices over the body surface of the human body. e major problem is that, due to the human body's lossy tissues, the performance of the antenna changes from the free space of operation. It is very important to check the performance of the antenna on the human body in order to check its suitability for the body area network. e antenna for WBAN applications also needs to be less sensitive to the human body.
Producing antenna arrays brings many advantages compared to a single-element antenna, such as increasing gain, creating wider bandwidth, neutralizing interference from a specific direction, increasing the signal-to-noise ratio (SNR), and precisely detecting the direction of the received signal. Several researchers proposed various series-fed array designs to increase their gain and bandwidth. [11][12][13][14]. At the joining point of the feed line and the patches, discontinuities in the junction cause phase shifts in the radiators. Moreover, the impedance matching technique becomes mandatory for the higher-order arrays. [15] To avoid this challenge, many researchers kept their arrays smaller. Besides this, smaller microstrip arrays mostly solve the gain and bandwidth issues that occur in a single-element antenna.
A  [19] with 68.3% efficiency at 60 GHz, which was a 4 × 4 element design. In [20][21][22], authors have proposed different broadband antennas with good performance parameters, but the antennas work at lower frequency bands. e antennas were designed on Rogers RT/Duroid with a relative permittivity of 2.2.
e antennas presented in [20][21][22] were not proposed for wireless body area networks. A textile-based wearable 60 GHz mmWave antenna for body-centric communication is proposed in [23]. First, the antenna was designed on a 100-polyester substrate, and then, it was designed on different other textile substrates. e antenna has an 11.63 GHz impedance bandwidth and shows a gain of 5.96 dBi. e antenna is a single element, and it has an overall length and width of 12.2 × 12 mm. A cotton-based mmWave array antenna for body-centric communications is presented in [24]. e antenna has an impedance bandwidth of 10 GHz and shows very low radiation efficiency. e overall size of the antenna proposed in [24] is larger.
In this paper, a design for a very small-sized mmWave antenna is presented with only a 3.96 mm long patch radiator, operating between 45 GHz and 62.5 GHz. Parallel-fed antenna arrays were designed with this element in several dimensions, such as 1 × 2, 1 × 4, and 2 × 2, and their simulated results were compared with each other. A human 2 International Journal of Antennas and Propagation body torso was created to check these array antennas' human performance on body performance. Several distances from the human body were also taken into account while comparing the simulated results. ough these results were computer simulation-based, simulations done in the CST environment were verified and found adequate through numerous research worldwide. Finally, a conclusion was drawn based on these results.

Antenna Design
e antenna was designed and tuned to operate in the 60 GHz range within the V-band of millimeter waves. A Microsoft Windows-based popular simulation software named "Computer Simulation Technique (CST) Microwave Studio Suite" was used to design the antenna and simulate it to check the desired parameters. e antenna, though having a similar look to the earlier design presented in [25], is actually a miniaturized version, which was done by reducing the size in all dimensions. It was a challenge to keep the design operational at the same desired frequencies while making it smaller and keeping the similar look. erefore, further tuning and parametric studies were done, and many structural dimensions were changed. In particular, the substrate part was completely restructured. Figure 1 shows the antenna model from different perspectives. e total antenna with the substrate is 5.12 mm long and 3.28 mm wide with a quasi-self-complementary design. At 60 GHz, the wavelength of the proposed antenna is 5 mm. e electrical size of the overall substrate length and width of this antenna is 0.97λ and 1.52λ, respectively. e antenna consists of multiple layers with a sandwiched air gap layer within the substrate. e material used as the substrate is Rogers RT5880, with a permittivity of 2.2 and a loss tangent of 0.001. e substrate, which is in total 0.306 mm thick, includes the 0.18 mm air gap layer between the 0.063 mm walls of RT5880 on both sides.
is carefully equalized thickness of the air gap was introduced to increase the antenna gain. On the top of the substrate, the radiator patch is located, which is the main antenna element for the 60 GHz operating frequency. By varying the size of this top radiating element, the operating frequency and the impedance bandwidth can be changed for this antenna. e radiator part alone is only 3.96 mm long and 1.54 mm wide. e length and width of the radiating element have an electrical size of 1.26λ and 3.24λ, respectively. e patch part is made of Perfect Electric Conductor (PEC) material. e thickness is kept at 0.035 mm. Below the substrate, the lowest layer is the ground, with a thickness of 0.035 mm, the same thickness as the radiator. is ground plane at the backside of the antenna usually helps to reduce the effects on human body tissues. Annealed copper was used for the ground with a conductivity of 5.8 × 107 S/m. Including all the layers with the air gap, the total thickness of the antenna is 0.376 mm.
e radiator patch has a feed of 2.01 mm long and 0.53 mm wide, connecting the waveguide port on the bottom and the main radiator patch. e patch is a combination of a half-elliptical part at the bottom and a half-triangular part at the top, with a 0.041 mm wide isosceles triangular slot in the middle. e slot is introduced to the top radiating element for impedance matching and wide bandwidth purposes. e equal sides of the triangle are 1.37 mm long, and the base is 1.23 mm. A 0.82 mm long and 0.37 mm wide rectangular brick is placed in the middle to tune the resonant frequency, with a pair of 0.082 mm slots on top and below it. e tip of the triangular part is sawed to make a 0.34 mm horizontal edge.
e detailed dimensions of the radiator patch are shown in Figure 2(a). On the other side of the substrate, the ground design is combined with a circular, triangular, and rectangular cutout. is helps to achieve a larger bandwidth of the antenna. e center of the circle slot is horizontally aligned in the middle of the substrate but vertically slightly shifted upward. e triangular slot was extruded and joined with the circular slot. e existing arms are defined in Figure 2(b) by 1.29 mm and 1.66 mm.
In Figure 2(c), the thickness of the different layers of the antenna is given from a side view, where the rightmost layer is the ground and the leftmost is the radiator patch. e air gap between the substrate and the air gap, which is 0.18 mm thick, is represented by the wider layer in the middle. Table 1 shows the different layers of the antenna with their corresponding thicknesses, materials used, and their relative permittivity. Figure 3, displayed above, the return loss (RL) curve of the antenna is presented. It shows the operating zone of the antenna around the 60 GHz frequency range, with a bandwidth of 17.8 GHz. ough the resonant frequency is slightly below 60 GHz, it has quite comfortably achieved an RL value below-24 dB at the desired frequency and has the upper cut-off at 65.562 GHz.

VSWR.
e curve below in Figure 4 shows the voltage standing wave ratio (VSWR) of the antenna. At 60 GHz, the VSWR value was 1.128, which is very close to unity. e voltage standing wave ratio graph shown in Figure 4 displays a consistent curve related to the return loss curve of the antenna. Figure 5, a 3D radiation pattern at 60 GHz is presented for the designed antenna, which shows a maximum gain of 4.01 dB in the direction along the XY plane, inclined to the X-axis. e 3D radiation pattern of this antenna looks very close to the omnidirectional pattern. A slight distortion is noticed in the radiation pattern, which is for the antenna structure and very high operating frequency. Figure 6, the radiation patterns of the antenna from a two-dimensional view are presented for the XY plane and YZ planes at 60 GHz frequency in Figures   International Journal of Antennas and Propagation 3.5. Surface Current. Surface current distribution on the ground and radiator patch is presented in Figure 7 when the antenna is excited from the connected waveguide port 1.

2D Radiation Patterns. In
Maximum current distribution is noticed over the top radiating element and near the feed of the antenna. Here, Table 2 shows the other parameters of the antenna. e antenna, having a bandwidth of 17.8 GHz, achieved a maximum gain of 4.01 dB with a radiation efficiency of 99.98%.

1 × 2 Array Design.
In order to improve the performance, this antenna was designed again as an array structure. A 1 × 2 array was designed with the earlier presented antenna with tapered lines and a parallel feeding technique (as shown in Figure 8). e substrate is 10.49 mm long horizontally and 7.17 mm vertically. Tapered feeds were made half the width gradually from the element's feeds, each with a 100-ohm λ/4 line to divide the power equally and connected to the 50-ohm feed with the width of fw. e antenna elements were kept a λ g distance apart; the technique followed from the [26] to enhance radiating gains, where λ g is the guided wavelength: where λ f /λ c is the ratio between the wavelength in free space and the cutoff wavelength. Or, simplifying more, where λ f is the wavelength of the EM wave at free space, and f c and f are the cutoff and operating frequencies, respectively. For a substrate with a dielectric constant of εr,      International Journal of Antennas and Propagation Figure 9 shows the backside of the array where the element grounds were aligned according to the radiator patches on the front. As the waveguide port was on the bottom of the antenna, the grounds were connected to the port with a 1.00 × 1.02 mm rectangular connecter made with PEC. e connector is the same height as the ground. e same number of free spaces was kept outside the elements on both the ground and patches. Figure 10 shows the 3D perspective views of the antenna. On the left, the two-element parallel-fed array antenna's front view is shown, while the back view is presented in the right figure.
e simulation results for the 1 × 2 design are given below.

Return Loss.
e return loss response of the 1 × 2 array is shown above in Figure 11. e curve shows a wide bandwidth with an RL value of around 18.5 dB at 60 GHz. Figure 12 shows the VSWR curve of the 1 × 2 antenna array. e voltage standing wave ratio (VSWR) was found to be 1.27 at the desired frequency within the mmWave spectrum range. Figure 13, the surface current distribution is shown when the antenna is excited with the waveguide port 1. e maximum surface current recorded was 409 A/m in the radiator patch area.

3D Radiation Pattern.
e 3D radiation pattern of the 1 × 2 antenna is displayed in Figure 14, where the maximum gain was recorded in the main lobe's direction. e recorded maximum gain was 5.01 dB along the YZ plane. e 3D radiation pattern for the 1 × 2 array for the antenna looks distorted, which is due to the effect of the array element. Figure 15 shows the two-dimensional radiation patterns of the antenna on both the XY and YZ planes. In Figure 15(a), the radiation pattern is shown on the XY plane where the sidelobe levels (SLLs) are lower than the main lobe. e main lobe magnitude was recorded at 2.42 dB with an angular width of 71.1 dB and the sidelobe level was −1.7 dB. e YZ plane exhibits an omnidirectional pattern. is 1 × 2 array structure shows improved gain compared to the single-element antenna. Figure 16 shows the 1 × 4 element array design with the proposed antenna. is was designed with the view of improving the performance of the antenna further. Elements were separated with λ g distance, as shown in Figure 17(a). ick feeds connecting the middle elements are 2.63 mm long. Elements were connected horizontally.

1 × 4 Array Design.
is parallel-fed design is connected with the same width 50ohm feed line as the original element, with a waveguide port on the bottom.
On the other hand, grounds were parallel connected to the port with similar rectangular PEC connectors as the 1 × 2 array, as shown in Figure 17(b). e ground elements were aligned with the radiator patches on the front. Grounds were placed vertically in the middle of the substrate, keeping 1.024 mm gaps on top and bottom. Other detailed measurements are given in Figure 17. e simulated parameters of the 1 × 2 array are given in the following sections.

Return Loss.
e return loss response of the 1 × 4 array is presented in Figure 18. e RL value at the desired frequency is slightly higher than the value found on the 1 × 2 array but still under -16 dB. It seems there are no major benefits added to the RL curve by making it a 1 × 4 in terms of bandwidth and the return loss value. Figure 19 shows the VSWR curve of the 1 × 2 antenna array. e voltage standing wave ratio (VSWR) was recorded as 1.3295 at 60 GHz. e VSWR curve similarly responded to the frequency responses found in the return loss for the 1 × 4 array.

Surface Current.
e surface current distribution of the 1 × 4 array is given in Figure 20 when the antenna is excited with the waveguide port 1. e maximum surface current recorded was 304 A/m near the rectangular brick in the radiator patch area. Figure 21, the 3D radiation pattern of the 1 × 4 antenna is shown where the maximum gain was recorded in the main lobe's direction along the YZ plane. e maximum gain recorded was 5.31 dB. erefore, the major difference was found to be in the radiation pattern in terms of gain and better directivity. e 3D radiation pattern of the 1 × 4 antenna looks like a better shape than the 1 × 2 antenna array structure. Figure 22 presents the 2D radiation patterns of the antenna on the XY (in Figure (a)) and YZ (in Figure (b)). In the XY plane's pattern, the International Journal of Antennas and Propagation sidelobe levels (SLLs) are lower than the main lobe, as in the previous array. e main lobe magnitude was recorded at 2.41 dB at 22°with an angular width of 60.3 dB, and the sidelobe level was -1 dB. e main lobe magnitude on the YZ plane was recorded at around 5 dB in a direction of 53°.

2 × 2 Array
Design. Another four-element array was designed with the antenna in a square matrix arrangement (2 × 2 array) to check the efficiency of the four-element array performance with different arrangements. e array was also constructed in a parallel-fed design. Figure 23 shows the 3D view of the 2 × 2 array with the antenna. Here, two 1 × 2 elements were connected with a 100 thin feed line. in transmission line feed width can be calculated using e simulation results for the 2 × 2 antenna array are given below. Figure 24 shows the dimensions of the different parts of the antenna. e total antenna array dimension in the front shown in Figure 24 Figure 25 shows the return loss response of the 2 × 2 antenna array. e return loss (RL) curve for the 2 × 2 antenna array shows a better result than the previous combinations shown earlier. At 60 GHz, the RL value was found to be below -15 dB, but the bandwidth is much wider. Figure 26 shows the VSWR curve of the 2 × 2 antenna array. e maximum surface current was recorded at 346 A/m around the rectangular shape in the middle of the patch antenna elements. Figure 26 visualizes the surface current distribution of the 2 × 2 arrangement of the antenna array.

Surface Current.
e maximum surface current was recorded at 346 A/m around the rectangular shape in the middle of the patch antenna elements. Figure 27 visualizes the surface current distribution of the 2 × 2 arrangement of the antenna array.

3D Radiation Pattern.
e 3D radiation pattern of the 2 × 2 array design is shown in Figure 28, where the maximum gain was recorded at 6.65 dB, which is a better increment than the other two array arrangements. Figure 29 shows the 2D radiation patterns of the 2 × 2 antenna array arrangements. On the XY plane, in Figure 29(a), the radiation pattern exhibits around 6 dB of gain in the direction of 30° e angular width of the main lobe is 14.1°and the sidelobe level is -0.5 dB. is antenna shows the highest gain for the 2 × 2 array design structure.

Comparison.
After simulating different antenna parameters for different array arrangements, the results are compared in the above figures. In Figure 30, return losses were compared for all three array designs with the singleelement antenna. In terms of RL values, all the designs show satisfactory return loss values at 60 GHz frequency. Bandwidth has significantly increased for the 2 × 2 antenna array.

International Journal of Antennas and Propagation
In the next comparison shown in Figure 31, gains gradually increased for 1 × 1, 1 × 2, 1 × 4, and 2 × 2 arrangements, which is visible in Table 3. e radiation efficiency of the 2 × 2 was found to be more optimistic compared to other arrangements. erefore, the 2 × 2 antenna array shows the most promising results through the simulations.
In Table 3, radiation and total efficiency are significantly better than any other array arrangement. Radiation efficiency was 80.39% for the 2 × 2 array. Similarly, the bandwidth and gain for the array are also much higher than the others. e gain of the array was 6.65 dB, while the single element was 4.01 dB. In terms of parameter improvement, the 2 × 2 array buildup was a successful attempt.

On-Body Performance Test
To check the eligibility of the antenna in the wireless body area network (WBAN), it is necessary to check the effectiveness of the antenna in close proximity to a human body.
ough the penetration depth of human skin by a 60 GHz millimeter wave is just a half millimeter [27], the presence of the human body greatly affects the antenna parameters like bandwidth, efficiency, return loss value, gain, and radiation patterns. Due to the capacitive nature of the human body, these parameters fluctuate [26].
For this reason, a three-dimensional human body model was made, consisting of the three outermost layers of the human body: skin, fat, and muscle (see Figure 32). e    model was built using the CST Microwave Studio Suite, including all the necessary electromagnetic properties of these body layers. First, a 1.3 mm thick skin layer was created with its specs, i.e., permittivity, conductivity, concentrations, etc. [28]. Similarly, a fat layer was created with a thickness of 3 mm adjacent to it. e innermost of these, the muscle layer, was also created and attached to it sequentially. For simulation purposes, 3 mm of the depth of the muscle layer was taken. e whole model is 7.3 mm thick. To achieve a more accurate result, the created human body model's length and width were kept greater than the value of 4. e created multiple array designs were then tested on the human body model by placing the antenna in four gradually increasing equidistant places, as shown in Figure  33, and resimulating all the parameters. Finally, the results are compared and commented upon the simulated results.

International Journal of Antennas and Propagation
For the single element, 1 × 2 and 2 × 2, the dimensions were kept the same as for the model, which was 16 mm × 19 mm (given in Figure 34(a)). To ensure radiation coverage for the 1 × 4 array's elongated horizontal structure, the dimension was modified to 28 mm × 19 mm (Figure 34(b)).
During the distance-based study, the antennas with different array arrangements were placed at four equal gap distances, starting from 2 mm to 8 mm, for each array, and the results were simulated to check and compare with each       other and the free space parameters. In Figure 34, how the antennas were placed at different distances is shown.

On-Body Test for Single Antenna (1x1 Design).
e solo antenna (1 × 1) was placed at 4 different distances: 2 mm, 4 mm, 6 mm, and 8 mm apart from the created human body model, and simulated the results one by one. e on-body test setup is shown in Figure 35.
All the distances provided similar outcomes in the return loss comparison. In the graph displayed in Figure 36, for the different distances, return losses increased as the distance decreased. e lowest return loss value was found for the 8 mm distance, which was -27.937 dB, while the highest was -18.81 dB for 2 mm. Due to the presence of the human body, there is a slight frequency shift noticed when the antenna is placed at different distances from the human body model.
In Figure 37, voltage standing wave ratios for different distances are compared. e changes in the voltage standing wave ratio are similarly relatable to the s11 curve, which was shown in Figure 36. Just like in the return loss comparison plot, the best value was recorded at 8 mm, which was 1.08.   Figure 32: 3D Human body model.

International Journal of Antennas and Propagation
A comparison of the radiation patterns of the solo element antenna for different distances is presented in Figure 38. Except for some gain value fluctuations, there is no discernible change in the patterns found on the XY plane (Figure 37(a)). On the YZ plane, changes are visible in the patterns for the presence of the human body compared to the free space pattern.
In Table 4, other parameters were compared. ere were no significant changes in the bandwidth for the presence of the human body or varying its distances. For the rest of the parameters, the results degrade when the antenna is very close to the human body (2 mm) and improve gradually when the distance is increased.

On-Body Test for 1 × 2 Array Antenna.
e 1 × 2 antenna array was placed at the selected distances of 2 mm, 4 mm, 6 mm, and 8 mm apart from the 3D human body model. After that, all the results were simulated again for various distances and compared. e on-body test setup for the 1 × 2 antenna array is shown in Figure 39.  International Journal of Antennas and Propagation In Figure 40, the return losses were compared for the different distances. No significant changes are visible in terms of RL values or bandwidths compared to the free space and on-body results. e minimum return loss value was recorded at -14.32 for an 8 mm distance.
In the VSWR curve shown in Figure 41, the results are very close to unity. All the values fluctuated between 1.38 and 1.48, while the free space value was 1.27. In the distancebased comparisons, the best result was found for 4 mm, which was 1.38. e results of the radiation pattern comparison on the XY and YZ planes shown in Figure 42 appear to be very stable, even when the array antenna is placed very close to the human body on both planes, with the exception of some minor gain value shifts.
In Table 5, as in previous simulations, there are no major changes in the bandwidth and return loss values for the distance varied body results. e gain and the radiation efficiencies fall a bit for the array antenna when it is placed 2 mm away from the human body model. As the distance increases, these parameters improve gradually, which is quite expected.

On-Body Test for 1 × 4 Array Antenna.
e designed 1 × 4 array antenna was positioned at the preselected four different distances from the 3D human body model. en, all the results were simulated again for various distances and compared. e on-body simulation setup for the 1 × 4 antenna array is given in Figure 43.     International Journal of Antennas and Propagation e return losses were compared for the different distances in Figure 44. As with earlier results, no significant changes are visible in terms of RL value or bandwidth compared to the free space and on-body results. e minimum return loss value was found for the 8 mm distance, which was -27.937 dB, while the minimum was -18.81 dB for 2 mm.
A slight shift in the resonant frequencies is visible. In the VSWR curve shown in Figure 45, results were oscillating around 1.5. 2D radiation patterns were compared on the XY and YZ planes presented in Figure 46. In both planes, results were very consistent in both the free space and on-body conditions. Distances affected the gain values slightly, but the patterns were very similar. Table 6 is presented where the gain and the efficiency act like the previously calculated simulation results; values become satisfactory when the distance is greater than 2 mm.
In Table 6, the antenna parameters except the gain for 2 mm distance were a bit degraded, but for other distances, all the other parameters like return losses, gains, or efficiencies improved gradually as the antenna to body gap increased.

On-Body Test for 2 × 2 Array Antenna.
e 2 × 2 arrangement of the array antenna was placed similarly at four different distances (2 mm, 4 mm, 6 mm, and 8 mm) and resimulated the results. Comparisons are given below. Figure 47 shows the on-body simulation setup for the 2 × 2 antenna array.
In Figure 48, the 2 × 2 array had a very wide bandwidth while tested in free space, which was also reflected in the onbody simulation, though the bandwidth was slightly reduced. Rather than this, the results remain much more stable for different distances. e minimum return loss was recorded at 4 mm, which was around −14.319 dB. Figure 49 shows the voltage standing wave ratio plots comparison for the array antenna's placement at different distances. It shows the VSWR values stayed around 1.5 in different scenarios too. Just like the s11 comparison curve displayed in Figure 48, the antenna showed the best results when placed at a distance of 4 mm. e maximum VSWR value was found to be 1.55 when the antenna was placed closest to the human body model.
In the following figure for the 2D radiation patterns (Figure 50), the array exhibited very consistent results across all the scenarios. Figure 50(a) shows the radiation pattern in the XY plane. Radiation patterns in the YZ plane shown in Figure 50(b) were compared, where the main lobe directions were shifted for different distances. Table 7 presents other parameter comparisons. In terms of gain and radiation efficiency, for the 2 mm distance, only the gain appears to be higher. However, for stable performance with better values, a 4 mm distance seems more appropriate. Overall, the 2 × 2 array antenna performs satisfactorily. Table 8 shows a comparison of the proposed antenna in this paper with the published articles. From Table 8, it is noted that the antennas presented in [20][21][22] are good designs with good performance, but they are proposed for the lower frequency band. A single-element antenna     International Journal of Antennas and Propagation working at 60 GHz was proposed for body-centric communication [23]. e antenna is designed on a 100% polyester substrate and it has a 11.632 GHz impedance bandwidth. However, this antenna [23] is not an array antenna, and the overall size of the antenna is bigger compared to the proposed antenna in this study. e antenna proposed in [23] shows lower gain and efficiency compared with the proposed design in this study. In [24], an array antenna is proposed for 60 GHz body-centric communication.
e antenna is designed with a cotton substrate [24]. e proposed design in [24] is bigger in size and has low efficiency. e proposed design antenna in this paper has the smallest size with an impedance bandwidth of more than 27.43 GHz. is proposed design also has very good gain and efficiency. e gain and efficiency vary for different array settings. is antenna construction is very new, and it is a small array antenna designed for bodycentric communications (BCCs).

Conclusion
e proposed miniaturized millimeter wave antenna performed satisfactorily with high efficiency near the 60 GHz frequency range (V-band) throughout the different simulations done. e antenna is quite suitable for tiny electronic devices and gadgets because of its smaller dimension and thinner and lighter structure, especially for the wireless body area network (WBAN). For gain and other performance enhancement purposes, several arrays were made with different orders and comparison yielded the most suitable and stable results with the 2 × 2 parallel-fed arrangement. In the human on-body test, the array with 2 × 2 elements showed much better stable performance with sufficient gain, directivity, and efficiency. When the proposed antenna with a 2 × 2 element array is placed on the human body model, it demonstrates 9.71 dBi gain and 57.32% radiation efficiency. In the future, this antenna can be optimized to work in other frequency bands of operation. In addition, a multiple-input and multiple-output (MIMO) algorithm can also be applied to this design [29].

Data Availability
e data used to support the findings of this study are freely available at http://niremf.ifac.cnr.it/tissprop/

Conflicts of Interest
e authors declare that they have no conflicts of interest to report regarding the present study.

24
International Journal of Antennas and Propagation