Design of an On-Glass 5G Monopole Antenna for a Vehicle Window Glass

In this paper, we propose a design for an on-glass 5G monopole antenna for a vehicle window glass. The proposed antenna consists of a monopole resonator, an inductive line, and a co-planar waveguide (CPW), which can adjust the phase so that the current in each resonator is close to in phase. Therefore, although the vehicle window glass has a high dielectric loss, the proposed monopole with an inductive line can obtain an antenna gain that is suitable for applying to vehicle 5G communications. To verify the antenna characteristics, the reflection coefficients and the radiation patterns are measured in a full anechoic chamber. The results demonstrate that the proposed on-glass antenna is suitable for applying in vehicle 5G communications.


I. INTRODUCTION
Recently, the demand has been progressively increasing for a vehicle-to-everything (V2X) communication technology to improve the level of autonomous driving technology [1]− [4]. V2X typically encompasses several specific technologies such as vehicle-to-vehicle (V2V), vehicle-to-pedestrian (V2P), vehicle-to-infrastructure (V2I), and vehicle-to-network (V2N), which can improve the safety and efficiency of autonomous driving. However, due to limited data throughput and latency issues, previous communication technologies are not suitable for V2X. Notably, 5G communication systems that use a millimeter wave (mmWave) have numerous advantages, such as low latency, high speed, and high capacity. Therefore, attempts to apply 5G communication techniques in autonomous vehicles have been increasing [5]− [7]. To apply 5G techniques to an autonomous vehicle, a high-gain 5G array antenna must be mounted on a small area of the vehicle. To address concerns regarding the appearance and air resistance of the vehicle, a shark-fin is generally used on the roof of the automobile for various vehicle antennas [8]− [10]. However, since a number of antennas (e.g., DMB, GPS, and LTE) are already built-in in the shark-fin housing, there is not enough space to insert an additional high-gain 5G array antenna without performance degradation from mutual coupling effect with adjacent antennas. To reduce the mutual coupling characteristics, various techniques such as the use of a cavity wall [11], a meta-surface [12], and a resonator [13] have been introduced. Nevertheless, the space is still not enough to insert these isolators in the shark-fin housing. In comparison, a glass antenna technique that prints an antenna pattern on the vehicle window has several significant advantages. First, the glass antenna does not require additional mounting space, which can preserve the appearance of the vehicles. Second, printing the antenna on the window glass prevents physical and electrical interferences (blockage or mutual coupling) with other vehicle antennas. Finally, the glass antenna can be easily made within the manufacturing process of the vehicle window glass. For these reasons, a glass antenna can be considered as a noteworthy candidate antenna type for applying in vehicle wireless communications, although thick glass windows with high dielectric loss characteristics does not help the performance of the 5G antenna at all. However, the previous studies on vehicle window glass antennas are limited in low frequency bands such as AM/FM, DMB, LTE, and 5G sub-6 band [14]- [16]. Although some studies on glass antennas have been attempted at a high frequency in the mmWave band, they were conducted on very thin glass, less than 0.5 mm [17]- [19]. Similarly, printed 5G antennas of general types such as Vivaldi antennas [20], slot antennas [21], and dipole antennas [22] do not consider glass substrates with high dielectric loss characteristics. In reality, the laminated glass for vehicle windshields with electrically very thick thickness has a high dielectric loss in the 5G mmWave band [23]− [25]. Therefore, in-depth research on a 5G antenna design for electrically thick window glass is required.
In this paper, we propose a design for an on-glass 5G monopole antenna for vehicle window glass. The proposed antenna consists of a monopole resonator, an inductive line, and a co-planar waveguide (CPW). The inductive line connecting the monopole resonators can adjust the phase so the current in each resonator is close to the in-phase. Therefore, although the vehicle window glass has a high dielectric loss, the proposed monopole with an inductive line can obtain an antenna gain that is suitable for applying to vehicle 5G communications. The monopole antenna is fed through a CPW transmission line. Herein, the monopole array and the CPW use only a single layer of vehicle window; therefore, it can be fabricated through a simple manufacturing process. Taking the smooth surface of the vehicle window glass into consideration, the conducting antenna body is attached using a thermosetting adhesive. To verify the antenna characteristics, the reflection coefficients and the radiation patterns are measured in a full anechoic chamber. The results demonstrate that the proposed on-glass antenna is suitable for applying in vehicle 5G communications.  Figure 1 illustrates the geometry of the on-glass 5G monopole antenna for vehicle window glass. The proposed antenna is attached to the side of the vehicle window glass, as shown in Figure 1(a). This window glass has a high dielectric loss characteristic (r = 6.95, tan = 0.05) with an electrically thick thickness of 3.2 mm. Therefore, it is difficult to obtain a suitable antenna gain with a 1/4 wavelength conventional printed antenna. To improve the antenna gain while simultaneously considering the high loss characteristics of the vehicle window glass, the proposed antenna consists of 4ⅹ1 monopole resonators with width and length of w and l, respectively. The monopole resonators are connected through the inductive line with width and length of wi and li, respectively. The inductive line can adjust the phase of each resonator, making the surface current in each resonator close to the in-phase. Herein, the total radiation pattern of the 41 monopole resonators with the inductive lines can be expressed as follows:

II. DESIGN OF THE PROPOSED ANTENNA.
where Fn is the radiation pattern of the nth single resonator (n = 1, 2, 3, 4), k is the wave number, and d is the array spacing. In addition, opt is the phase delay by the optimized inductive line that can makes the surface current in each resonator close to the in-phase. In an ideal situation, the 41 monopole resonators with perfect in-phase state for each resonator have a higher gain of 6 dB than a single monopole resonator. Moreover, to obtain a higher antenna gain, the 41 monopole can repeatedly arrange in the x-axis direction. For example, the radiation pattern of the 44 monopole array expresses as the summation of X41n as follows: Likewise, the 44 monopole array gain is higher of 6 dB than the 41 monopole array in an ideal situation. Therefore, although a vehicle window has a high dielectric loss characteristic, the proposed antenna can obtain the antenna gain suitable for applying 5G communications The proposed antenna is linearly polarized, which is widely employed in vehicle 5G communications because it has some advantages such as ease of an antenna design and fabrication [26]- [29]. An input signal is fed to the proposed antenna through the CPW line with a width of wf and a gap of g. Since the monopole resonators and the CPW use only a single layer of vehicle window, it can be fabricated through a simple manufacturing process. The detailed design parameters for maximizing the bore-sight gain are obtained with the FEKO electromagnetic (EM) simulator [ Figure 2 presents a photograph of the fabricated 41 monopole antenna. In general, since vehicle windows have very smooth surfaces, it is difficult to deposit a copper layer on the glass substrate. Therefore, the glass antenna cannot be fabricated using the usual method for printed antennas. To print the conducting antenna body on the vehicle window, a silk screen method is used in general. However, this method has poor manufacturing tolerance, so it is not suitable for fabricating 5G antennas with small antenna patterns. To solve these manufacturing problems, a thermosetting adhesive is used to attach a copper film on the vehicle window glass. The thermosetting adhesive can maintain adhesion between the glass and the antenna body when the antenna pattern is heated for the soldering process. At the next stage, the vehicle window with an attached copper film is etched, which makes it possible to form a small antenna pattern despite the fact that the vehicle window glass has a smooth surface. A part of the CPW line is connected to a Ktype (2.92 mm) port that can operate the antenna with a low insertion loss at the 5G mmWave band. To verify the fabricated antenna, the reflection coefficients and the radiation patterns are measured in a full anechoic chamber.  Figure 3 presents the reflection coefficients of the proposed antenna. The solid and dashed lines indicate the measured and the simulated reflection coefficients, respectively. The measured reflection coefficient maintains a value that is less than −10 dB in the observed frequency band, which is in good agreement with the simulated results. Therefore, the proposed antenna can be used in the 5G mmWave band.   Figure 4 shows the 2-D radiation patterns of the proposed antenna on the zx-and zy-planes. The measured and simulated co-polarization gains (bore-sight direction) are 0 dBi and 0.4 dBi at 28 GHz. The gain of the proposed antenna is not higher than that of conventional monopole antennas, because the high dielectric loss of the commercial vehicle glass causes the decrement of the radiation efficiency. In fact, the proposed antenna has a low radiation efficiency of 11.1% at 28 GHz. Of course, if the proposed antenna is designed with a low loss substrate, then the antenna gain can be enhanced as much for the conventional monopoles. On the other hand, the cross-polarization gain (bore-sight direction) is −20.1 dBi, which means that the proposed antenna has linear polarization characteristics. In addition, the half power beamwidths (HPBWs) are 18 (zx-plane) and 27 (zy-plane), which shows good agreement in comparison to the simulated HPBWs of 14 (zx-plane) and 29 (zy-plane). To verify the feasibility, the antenna characteristics such as an operating frequency band, a maximum gain, antenna dimensions, and a substrate material of the proposed antenna are compared to those in other previous studies, as listed in Table 2.

III. VERIFICATION OF THE PROPOSED ANTENNA
To verify the operating principles of the proposed on-glass antenna, the 5G monopole antenna is analyzed from a circuit perspective. Figure 5 indicates an equivalent circuit model for the proposed 4ⅹ1 monopole antenna. The equivalent circuit model is obtained using a data fitting method. Herein, Rn, Ln, and Cn indicate the resistance, inductance, and capacitance components of the nth resonator (n = 1, 2, 3, 4), while Lf is the feed inductance. The mth inductive line from the source connecting the monopole resonator is specified as Lim (m = 1, 2, 3). The parameters of each lumped element are adjusted iteratively to minimize the impedance difference between the circuit model and the full EM simulation. The results are compared in Figure. 6. The detailed parameters of the lumped elements are listed in Table 3.         Figure 8 illustrates the geometry and photograph of the expanded 44 array configuration that can obtain enhanced bore-sight gain than the 41 monopole. For this configuration, the 41 monopole is repeatedly arranged in the x-axis direction, and the array spacing is determined to be 8.4 mm. The 44 array antenna is manufactured using the same methods described in Figure 2. The fabricated 44 array antenna is measured in a full anechoic chamber to obtain the radiation patterns of the array antenna.  Figure 9 shows the 2-D radiation pattern of the proposed 44 array antenna on the zx-plane. The solid and dashed lines indicate the measured and the simulated radiation patterns. The bore-sight gains of the measured and the simulated are 5.1 dBi and 3.9 dBi at 28 GHz, respectively. It shows a 5.1 dB higher gain than the 41 array configuration, which is good agreement with the theoretically expected result. In addition, the HPBW is 11.5, which shows good agreement with the simulated HPBW of 15.

IV. CONCLUSION
We investigated the design of the on-glass 5G monopole antenna for the vehicle window glass. The proposed antenna consisted of the monopole resonator, the inductive line, and the CPW. The inductive line connecting the monopole resonators, could adjust the phase so that the current in each resonator was close to in phase. Therefore, although the vehicle window glass had the high dielectric loss, the proposed monopole with the inductive line could obtain the antenna gain suitable for applying to vehicle 5G communications. To verify the antenna characteristics, the reflection coefficients and the radiation patterns were measured in the full anechoic chamber. The measured reflection coefficient remained at less than −10 dB in the 5G mmWave band. In addition, when the array antenna was in the 44 configuration, the measured bore-sight gain and HBPW were 5.1 dBi and 11.5 at 28 GHz, respectively. To analyze the operating principle of the proposed on-glass 5G monopole antenna, the imaginary characteristics of impedance were examined for the equivalent circuit model and the full EM simulation model. In addition, the surface current phase variance of the monopole antenna was observed and compared both with and without the inductive line. The results demonstrated that the inductive lines were required to improve the gain by adjusting the surface current phase of the resonators. On the other hand, the proposed antenna had the small size but not optically transparent. To enhance the visibility, transparent conductors or mesh structures can be applied to 5G glass antenna in future work.