3.5 GHz Rectangular Patch Microstrip Antenna with Defected Ground Structure for 5G

This research has performed the design and implementation of microstrip antenna for fifth generation (5G) application, at frequency 3.5 GHz. The desired parameters are based on Huawei public policy position, Qualcomm public policy position, and Rel-15 3rd Generation Partnership Project (3GPP) article. Since microstrip antenna has narrow bandwidth, some modification are conducted, namely proximity coupled feeding and defected ground structure (DGS). The first stage is calculating the initial dimension of the antenna, finally the antenna is simulated and optimized. The simulation starts from simulating the initial dimension, then applying the proximity coupled feeding, after that employing the DGS until the desired antenna is achieved. The final stage is fabricate the antenna based on simulation then measure it. The measurement results show that the gain is increased to 6.6 dB, the bandwidth is reduced by 65.2 MHz, the Voltage Standing Wave Ratio (VSWR) and return loss are 1.31 and -17.436 dB.


INTRODUCTION
The fifth generation (5G) is the latest generation of cellular mobile communications. It succeeds the fourth generation (4G) with the Long Term Evolution (LTE) and Worldwide Interoperability for Microwave Access (WiMax), the third generation (3G) Universal Mobile Telecommunications System (UMTS) and second-generation (2G) Global System for Mobile communications (GSM) systems. 5G performance targets high data rate, reduced latency, energy saving, cost reduction, higher system capacity, and massive device connectivity. For instance, for the downlink (DL), the experienced data rate of up to 50 Mbps are expected outdoor and 1 Gbps indoor (5GLAN), and half of these values for the uplink (UL), with the target for user plane latency, should be 4 ms for UL, and 4 ms for DL (3GPP, 2019). As 5G still on progress, there is no fix regulation in Indonesia, expected to be announced in 2020 or 2021 (PERTIWI, 2018).
However, some institutions already have major parameters and specification requirement for this next-gen communication, such as 3 rd Generation Partnership Project (3GPP) that has already finished the first phase. Also, some big companies have already published the specification for their future devices that will use 5G technology later, such as Huawei and Qualcomm, as pioneers for 5G patents in east and west respectively (Qualcomm, 2017), (Huawei, 2017). Figure 1 shows the frequency aspect for the 5G technology according to the 3GPP Rel-15 on 5G New Radio (NR) of two frequency ranges which are FR1 being sub-6 GHz range (450 -6000 MHz) and FR2 being the millimeter wave (mmWave) range (24250 -52600 MHz) (3GPP, 2019).
Huawei also notes that, as the first steps, it highly recommend that the countries allocate 3300-3800 MHz or a portion of it and make it available for 5G with consistent timelines and regulatory frameworks, i.e., frequency arrangements and emission masks. It recommends that at least 100 MHz of contiguous bandwidth from this band allocated to each 5G network. The frequencies for 5G are divide into three groups, that is low frequencies, medium frequencies, and high frequencies (Huawei, 2017). The design of microstrip patch antenna is a singlelayer consist generally of four parts, i.e., the patch, ground plane, substrate, and feeding part. The physical size of a microstrip antenna is small. However, the electrical size measured in the wavelength is not so small. The most commonly employed microstrip antenna is a rectangular patch (Paul & Sultan, 2013).
Microstrip patch antennas have more advantages and better prospects compared to the conventional antennas, such as lighter in weight, low in volume, cost, profile, and smaller in dimension, as well as ease of fabrication and conformity. Moreover, the microstrip patch antennas can provide frequency agility, broad bandwidth, feedline flexibility and beam scanning omnidirectional patterning (Bisht, Saini, Prakash, & Nautiyal, 2014). There are also some disadvantages such as narrow bandwidth, low gain, generate unwanted radiation because of its unification technique and low efficiency. To overcome the disadvantages, we can use several approaches, such as changing the feedline and making a defected ground plane.

Calculation the Size of Patch
The formula used for designing the width of a patch antenna (W) are as ( The electrical length of a patch antenna is greater than the physical. This normalized extension in length is calculated using (Paul & Sultan, 2013) The actual length of a patch antenna (L) calculated using (Paul & Sultan, 2013) ɛ 2 (4)

Determining the Size of Feedline
The feeding channel used in this paper has the impedance value equal to 50 Ω calculated by 5.7961 ɛ .
where is the feedline width. From the Equation (1) -(5) the initial design of microstrip antenna single patch is shown by Table 1 as well as Figures 3 and 4. Table 1 shows the initial size of the antenna for the simulation. Figure 3 shows the substrate where the yellow area is the patch and the feedline. Figure 4 depicts the antenna ground plane and the antenna design.

Initial Simulation Results
The calculation results of the above design using Ansoft HFSS 15.0 shown in Table 2 and Figures 5 and 6. As shown in Table 2, some of the desired parameters have not been achieving yet. The bandwidth and working frequency, which are still far from the desired, also needs to improve the other parameters. Figure 5 shows that the bandwidth achieved from the simulation is 0.1392 GHz or 139.2 MHz when the return loss is -16.32 dB. Furthermore, the working frequency at -10 dB is nowhere near 3.4 -3.6 GHz. Figure 6 shows that the VSWR is 1.3603. It is required to modify the antenna dimension to achieve the desired parameters.

Changing Patch Size & Feedline Length
For optimization, there are several modifications to the antenna dimensions. The first step is modifying the patch size. Table 3 shows the effects of changing the dimension of the patch. According to Table 3, the working frequency is close to the desired parameters. The second step is modifying the length of the feedline, shown in Table 4.

Changing Substrate Size
From Table 4, the best length for the feedline is 17.5 mm. However, the working frequency is shifts a little. The next step is to reduce the overall dimension of the antenna, i.e., the ground plane and the substrate. The first dimension is 50 x 40 mm and then changed to 40 x 30 mm. The result is that it shifts the working frequency to 3327.3 -3495.5 MHz, reduces the bandwidth to 169.2 MHz, return loss is -13.3654 dB and VSWR is 1.55. It is because the dimension 40 x 30 mm leads the parameters changed too much. Then modify the dimension to 45 x 35 mm. Hence, Table 5 shows the results. With this modification, we achieve the desired parameters. However, the working frequency still shifts. Hence, we modify the dimension of the patch and the length of the feedline. Table  6 shows the effects in modifying the patch dimension and the length of the feedline with 45 x 35 mm substrate. We achieve the desired working frequency when the patch dimension is 18 x 26.2 mm, and the length of the feedline is 15.2 mm.

Applying DGS on Ground Plane
Implementing Defected Ground Structure (DGS) in microstrip patch results in improves the antenna bandwidth. Practically bandwidth obtained using DGS is up to 100 MHz Along with bandwidth, other parameters such as the uniform current distribution, beamwidth, return loss, reflection coefficient, VSWR, are also improved. The final step of the optimization is modifying the ground plane with DGS. Figure 7 shows the experimental DGS used in the simulation. It should note that Table 7 shows the effect of modifying the ground plane of the antenna.

Antenna Final Size and Simulation Results
Based on Table 7, the last result is satisfied with the desired parameters. With the working frequency close to the desired parameters, it achieves better VSWR and positive gain. The following Table 8 as well as Figure 8 and Figure 9 are complete specifications and simulation results of Rectangular Patch Microstrip Antenna with Proximity Coupled Feeding Technique and Defected Ground Structure. Table 8 shows the final size of the antenna design that we will fabricate. Furthermore, Figures 8 shows the simulation result of the final design, i.e., the antenna working frequency, return loss, and VSWR. Next, Figures 9 shows the radiation pattern and gain of the antenna.

Fabricated Antenna
Based on the simulation design using HFSS Ansoft 15.0 software, the fabricate antenna shown in Figures 10 and 11. We fabricate the antenna on a double layer printed circuit board (PCB) that consists of two PCB. The first PCB or substrate etching for the patch of the antenna. And, the second substrate for the feedline and the ground plane of the antenna. Figures 10 and 11 show the etching result. Then, combine the two substrates, in this case, the first and second substrate with duct tape. On the feedline area, a connector soldered. The difference between the simulation and the measurement results are occurred by some several factors as follows.
(1) The fabricated antenna shows that there is reduces in thickness.
There is a trail of sandpapering the substrates, hence the thickness of the antenna is reduced. The change of thickness affects the bandwidth of the antenna. In addition, (2) the fabricated antenna manufactured conventionally by etching a double layer printed circuit board (PCB). This method also can cause the thickness of the substrates to decrease. Moreover, (3) when the two substrates combined, there is little gap near the connector.
where P1 is the gain value of two identical horn antenna, P2 is the gain value when microstrip antenna tested as a receiver, and GReff is the gain value of the reference antenna. Referring to Equation (6) with P1 of -33 dB, the value of P2 of -29.51 dB and the reference antenna gain value of 10.1 dB, the antenna gain value is 6.61 dB. Table 9 shows the final simulation and measurement results. It shows the simulation and measurement results are not in accordance, but both still satisfy the desired parameters.

CONCLUSION
There are some differences between the simulation and the measurement results, but the parameters are considered well enough, as it is satisfied the desired parameters. It is because the bandwidth decreased by a significant amount. The results in this research are that the VSWR increased by 0.242, the impedance increased by 15.256 Ω, the gain increased by 3 dB, and the return loss changed to -17.436 dB. The VSWR, return loss, and gain are satisfie the desired parameters. Also, the bandwidth satisfies both the public position of Huawei and Qualcomm. Huawei advises that the bandwidth for 5G at least 100 MHz. While, Qualcomm advises that for mid-band 5G, the bandwidth is 150 MHz. Finally, the simulation and the measurement results show that the antenna is in accordance with the desired parameters. It expected that hopefully, the antenna is useful for 5G applications.