Comparative Analysis of a Super-Wideband Millimeter Wave Array Antenna for Body-Centric Communications

Department of Electrical Engineering, Faculty of Engineering, Jouf University, Sakaka 42421, Saudi Arabia Department of Electrical and Computer Engineering, North South University, Dhaka-1229, Bangladesh Centre for Telecommunication Research & Innovation (CETRI), Faculty of Electrical and Electronic Engineering Technology (FTKEE), Melaka (UTeM), Melaka 76100, Malaysia Department of Information Technology, College of Computers and Information Technology, Taif University, P.O. Box 11099, Taif 21944, Saudi Arabia Electrical Engineering Department, Faculty of Engineering, King Khalid University, Abha 62529, Saudi Arabia Electrical Engineering Department, Faculty of Engineering, Aswan University, Aswan 81528, Egypt Faculty of Computers and Information, South Valley University, Qena 83523, Egypt Applied College, Department of Technology, Taif University, P.O. Box 11099, Taif 21944, Saudi Arabia


Introduction
Demand for high data rates and the usage of smart devices has increased exponentially in the last few decades. To avoid interference and overburdening of current microwave systems, millimeter wave (mmWave) is fast becoming the preferred range of communications. e 60 GHz band of the mmWave has a bandwidth of more than 7 GHz and is suitable for short-ranged indoor applications. e band is unlicensed in different countries, which makes it very promising. High data rates of more than 7 Gbps can be achieved with low interference and a high level of security [1][2][3][4]. At 60 GHz, the atmospheric absorption is very high and, along with propagation losses, the range of mmWave is very short. On the other hand, this property gives high security as it becomes very difficult for other neighboring networks to interfere with signals. Until now, the 60 GHz band was exclusively available for noncommercial applications like the military and satellites [5]. e unlicensed 60 GHz band is gathering a lot of interest from researchers who are focused on developing systems such as healthcare and multimedia. In systems like health care, the nodes are connected around the human body. e networking between these nodes is known as a wide-body area network or WBAN. Antennas for WBANs must be tiny and flexible, and appropriate radiation patterns must be considered based on the position of the node. Furthermore, the human body's influence on the performance of the antenna should also be studied [6,7].
From link budget analysis, it can be concluded that a high gain antenna is a core requirement for the 60 GHz band after considering several factors such as distance, power constraint, and bandwidth. [8]. A disc-shaped antenna with 5.2 dB gain is proposed in [9]. A wearable textile antenna with 11.9 dBi gain for WBAN applications is presented by Chahat et al. [10]. A compact novel design and a comparative analysis of a 60 GHz antenna are presented in [11,12] for body-centric applications that achieve a maximum gain of 9.73 dB to 10 dB. To raise the gain even further, antennas can be arranged in linear or planar arrays. Zhan et al. presented two different 60 GHz patch arrays in [13]. e first design is a 4 × 1 linear array, and the second design is a 4 × 4 planar array. e linear array achieved a bandwidth of 10 GHz with a maximum gain of 12.5 dBi. In comparison, the planar array achieved a slightly lower bandwidth, but the maximum gain increased to 18 dBi. Two Yagi array designs with two and four elements are presented in [14]. e 4-element array achieved a bandwidth of 8.2 GHz with a maximum gain of 13.85 dBi and an efficiency of 95.4%. A 7 × 7 phased array antenna with a gain of 22.15 dBi is presented in [15]. Four different array configurations have been proposed in [16]. e maximum gain varied in the range of 12.8 to 15.6 dBi. Hong and Choi explored a 4-port array design under human phantom conditions in [17]. Wu et al. developed an E-shaped patch array for the 60 GHz band that achieved a maximum gain of 17.92 dBi [8]. e gain of a four-element patch array based on a cotton textile substrate was 12.6 dBi [18]. e same design was studied on a RT 5880 substrate in [19] with a 3 mm thick ground plane in order to achieve a low SAR value. Compared to a single patch, the array configuration achieved a 5 dBi higher maximum gain. Leduc and Zhadobov in their paper proposed three different feeding techniques for the four-element patch array to reduce radiation exposure to the human body [20]. Two different Yagi designs were presented in [21] in order to achieve a wide bandwidth and high gain. e conventional Yagi design achieved a high gain of 17 dB with a bandwidth of 6 GHz, while the SIW Yagi achieved a bandwidth of 7 GHz with a slightly lower gain of 15 dB. A T shape mmWave array antenna for 5 G system is proposed in [22]. e antenna is designed on a Roggers substrate with a dielectric constant of 2.3. It works at 28 GHz with an impedance bandwidth of 8 GHz.
e overall size of the antenna presented in [22] is 25 mm × 18.85 mm. A 60 GHz mmWave antenna designed on different textile substrates is presented in [23]. e antenna was proposed for wearable application, and its on-body performance was investigated. It was designed on a 100 percent polyester substrate at first and afterwards on several textile substrates. e antenna's overall dimensions are 12.2 mm by 12 mm, and it operates at 60 GHz with an 11.632 GHz impedance bandwidth [23]. A Q-slot antenna for on-body applications is given in [24]. It operates at 60 GHz. e antenna is printed on the substrate with an overall size of 12.9 mm × 14 mm. e antenna presented in [24] has an impedance bandwidth of 12.11 GHz.
In this paper, we are proposing three different array configurations of a novel microstrip patch antenna design. e aim is to achieve a super wide bandwidth and a high gain. As reported in [21], a compromise is needed between bandwidth and gain. Our goal is to achieve a bandwidth of more than 20 GHz with a sufficient amount of gain for short transmissions. e patch elements are arranged in 2, 3, and 4 array configurations. Each configuration is studied under human conditions by considering a three-layer phantom. e paper is divided into five sections. e three array designs are laid out in Section 2. e third section consists of free-space simulation results. Section four contains the findings of the on-body simulation, and Section 5 contains the conclusion. e design and simulations were done using CST Microwave Studio.

Antenna Design
Each single element of the three arrays consists of a semicircular disc patch. e ground plane has a novel shaped slot etched into it that is complementary to the radiating patch. All three antenna arrays are designed on a 1.5 mm thick FR-4 substrate which has a relative permittivity of 4.3. e radiating patches are made from 0.035 mm thick PEC. e material thicknesses are presented in Table 1.
e first of the three proposed array designs is a twoelement array (Array 2), presented in Figure 1. A 2.5 mm long, 0.7-width microstrip line feeds the patches, which are 9.5 mm apart. e ground plane is 2.5 mm apart. e detailed dimensions of Array 2 are marked in Figure 1, and the values are given in Table 2. e Array 3 design consists of three patches (Figure 2). e middle patch and the right side patch are facing each other and are the same distance apart as Array 2. e left side element is 8.1 mm away and is facing in the opposite direction to the middle patch. e detailed dimensions of Array 3 are given in Table 3. Similarly, for Array 4, one more element is added 8.1 mm away to the right side of Array 3 and is facing in the opposite direction. e dimensions are detailed in Figure 3 and Table 4. e antenna's wavelength at 60 GHz is 5 mm. e electrical size of the length and width of the substrate of a two-array antenna is 0.5 λ and 0.3 λ, respectively. e electrical size of the length and width of the substrate of a three-array antenna is 0.5λ and 0.19λ, respectively. Similarly, the electrical size of the length and width of the substrate of a four-array antenna is 0.5λ and 0.14λ, respectively.

Free-Space Simulation Results
All three array antennas achieved a very wide bandwidth. e return loss magnitude was less than10 dB for more than 20 GHz, which covers beyond the unlicensed 60 GHz band. At 60 GHz, the return loss magnitude improved as the number of antenna array elements increased. e return loss curves are depicted in Figures 4-6 for Array 2, Array 3, and Array 4, respectively. e comparison return loss curve is depicted in Figure 7. For all three arrays, the VSWR is less than the acceptable value of 1.5 over the whole simulated range. e individual VSWR is depicted in Figures 8-10, and the comparison is in Figure 11. e H-plane radiation pattern at 60 GHz for Array 2 is omnidirectional, while the E-plane demonstrates a grate pattern ( Figure 12). For Array 3, the patterns are more directional, as seen in Figure 13. Array 4 shows even more grated patterns ( Figure 14). e gap distance between the antenna elements is responsible for this grating pattern. e polar plot comparisons in Figure 15 and the 3D radiation patterns of the three designs are presented in Figure 16. e maximum gain is 3.44, 4.92, and 6.19 dBi for arrays 2, 3, and 4, respectively. e radiation efficiency is quite low at 41.51% for Array 2. Array 3 achieved     International Journal of Antennas and Propagation    Figure 17. e free-space simulation results are summarized in Table 5.

On-Body Simulation Results
Due to the absorption of electromagnetic waves, the antenna's performance is greatly affected by the human body. e radiation pattern gets distorted, and parameters such as efficiency, gain, and return loss are changed. A human torso is built to have the three outermost layers of a human body-skin, fat, and muscle-to replicate an on-body scenario ( Figure 18). Table 6 shows the thickness, relative permittivity, and conductivity for these layers at 60 GHz [22]. Depending on the array configuration, the size of the phantom is changed to accommodate the antenna design ( Figure 19 and Table 7). For every configuration, the thickness of the three layers is kept constant. For on-body analysis, all three antennas are placed at five different positions away from the phantom at 2 mm intervals ( Figure 20). e return loss curve of Array 2 shifted to the left at different distances from the phantom (Figure 21). e shape of the curve at different distances remained comparable to free space. A similar pattern is observed for the VSWR ( Figure 22). e radiation pattern shows reduced back radiation and grated lobes towards the phantom. e patterns are more grated in the E-plane, whereas the H-plane is comparable to the free space at all distances. e patterns are depicted in Figure 23, and the 3D radiation pattern in Figure 24. Maximum gain was highest at 6 mm away from the phantom. At every other distance, the gain was higher than free space, except at the closest distance from the phantom. e radiation efficiency and the total efficiency decreased by a significant margin, especially at the closest distance to the phantom. ese values improved as the gap distance increased gradually. e on-body performance details of Array 2 are summarized in Table 8.
At 8 mm away from the phantom, Array 3 had a similar return loss to free space. e resonant frequencies were closer to free space at this distance. e overall shape of the curves is very similar to the free space ( Figure 25). e corresponding VSWR is depicted in Figure 26. e maximum gain improved to more than 5 dBi at 6 mm away from the phantom. At greater distances, the gain was closer to 5 dBi. e radiation patterns at different distances for Array 3 are depicted in Figure 27 and the 3D pattern in Figure 28. e return loss curve of Array 4 reveals that the antenna was resonant extremely close to 60 GHz when it was 6 mm away from the phantom (Figure 29). e VSWR curve shows that at around 60 GHz, the value was very close to 1 (Figure 30). e radiation patterns depicted in Figures 31 and 32 show similarities with the free space patterns at greater distances from the phantom. e detailed on-body performance values are summarized in Tables 9 and 10 for Arrays 3 and 4, respectively.
Arrays 2, 3, and 4 achieved comparable return loss responses to the free space for all five distances (Figure 33). e return loss curves are slightly shifted to the left of the free space. e VSWR curves at 4 mm away from the phantom are compared in Figure 34. e radiation patterns of both the E and H planes show reduced back radiation with more gratified lobes. e patterns improved and resembled the free space patterns at 10 mm for all three arrays. e patterns for all three arrays at 4 mm are compared in Figure 35. e     [18]. e antenna is also larger in size in comparison to the antenna presented in this study. In [23], a polyester-based textile antenna is presented, which is not an array. e antenna presented in [24] is based on the FR-4 substrate and it is not an array antenna. e proposed antenna design in this study is an array antenna. It is more compact compared to other articles. e antenna in this study has a      Figure 18: Phantom consisting of three outer most layer of human body.    International Journal of Antennas and Propagation

18
International Journal of Antennas and Propagation

Conclusion
In this paper, we have presented three different arrays of a 60 GHz super-wideband antenna. Due to a lack of adequate facilities, no measurements were conducted, so we only provided simulated findings. e CST simulation results are reliable, and we expect any future measurements taken will be within the margin of error. Arrays are designed to achieve high gain, which is required to increase the range at 60 GHz. e gain increased as the number of elements increased, but not at the same rate as in other reported works. e maximum gain achieved by all three arrays is suitable for very shortrange communication. Efficiency was also affected, especially under on-body scenarios. Microstrip patch antennas usually have low efficiency, low bandwidth, and a broadside radiation pattern. Results presented in this paper confirm this, even though we managed to achieve high bandwidth. On-body applications require broadside radiation, which is achieved by these arrays. e antenna designs presented in this paper are very small, and thus careful fabrication is required to take performance measurements. e proposed antenna in this study is novel compared to the presented antennas in the open literature.
is structure of the antenna is new and is a compact array antenna for body-centric communications (BCCs). It is a more compact antenna compared to the array antennas proposed for BCCs in other articles. Although the size of this presented array antenna is smaller, it shows a huge bandwidth of more than 20 GHz. e gain of the antenna increases as the number of antenna arrays is increased. e presented antenna shows more than 6 dBi gain. is antenna shows very good on-body performance. is study and the findings in this paper are comprehensive.

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.   20 International Journal of Antennas and Propagation