Antenna element design

Figure 1 shows the exact structure of the antenna element. A 0.1 mm thick LCP dielectric substrate with a loss tangent of 0.002 and a relative permittivity of 2.9 is used in the suggested antenna. The front is a 50Ω feeder, and the background has an engraved H-shaped slot. The H-shaped slot is added to improve the impedance bandwidth of the antenna. The overall size of the element is 18 × 16.5 mm², and the simulation bandwidth covers 5.5–6.21 GHz. The optimized geometric parameters of the proposed antenna are as follows: LL = 16.5 mm, WW = 18 mm, H = 0.1 mm, Wc = 16 mm, Lf = 16 mm, Lk = 8.5 mm, M1 = 6 mm, M2 = 3 mm, Wf = 0.26 mm, Lc = 1 mm, G = 0.1 mm.

Figure 1. Antenna element structure diagram. (a) Top view (b) Back view.

The antenna element design process is given in Fig. 2. Antenna I is the first design step, the front is a 50Ω feed line, and the background has an engraved H-shaped slot, which can expand the bandwidth of the antenna. Antenna II is the second design step, the original H-slot is changed to a double H-slot for reducing the size of Antenna I, and the feed line is extended to both sides to form a T-shaped structure. Through miniaturization, the final antenna element’s total size has been lowered by 45%.

Figure 2. Design process of antenna element.

The S-parameters for both two antennas are presented in Fig. 3. Antenna I covers 5.31–6.45 GHz, and Antenna II covers 5.5–6.21 GHz. After miniaturization, the bandwidth of the antenna is shortened by 0.43 GHz, but the reduced bandwidth of Antenna II still covers the desired frequency band. As the size of Antenna II is smaller, so it is chosen as the unit of the antenna array.

Figure 3. S₁₁ of the two antennas.

In Fig. 4 the current distribution of Antenna I and Antenna II at 5.8 GHz is given. From the figure we can notice that the current is mainly concentrated around the slit and feedline. The antenna’s miniaturized design reduces the antenna’s size while increasing the current path, which ensures that the antenna resonates at 5.8 GHz. Figure 5 gives the relationship between the length (Lf) and the antenna’s resonance frequency. In the figure, when the Lf becomes larger, the antenna’s electrical length increases, and the resonant frequency is more biased toward the low frequency.

Figure 4. Current distribution diagrams. (a) Antenna I (b) Antenna II.

Figure 5. The influence of Lf on S_11.

Figure 6. Antenna array structure. (a) Top view (b) Back view.

Antenna array

Four antenna elements and the power divider are combined to form a quaternary antenna array, as shown in Fig. 6. The power divider is designed using the Chebyshev unequal division method to increase the antenna’s directivity. Overall antenna dimensions are 111 × 27 mm², and the spacing between adjacent units is 30 mm. Figure 7(a) displays the antenna array’s S₁₁ parameter, which covers 5.13–5.94 GHz. As the antenna resonance frequency will be shifted to the high frequency by adding AMC structure, so the antenna resonance frequency is set at 5.58 GHz. At 5.8 GHz, the realized gain and side lobe are 9.7 and −17.5 dB, respectively. The radiation pattern is shown in Fig. 7(b).

Figure 7. Antenna array performance. (a) S₁₁ parameter (b) Radiation pattern.

Figure 8. The proposed AMC unit. (a) AMC unit model (b) Equivalent circuit model.

AMC unit

To lower the radiation toward the human body as well as to increase the FBR of the antenna, an AMC structure reflector is loaded on the back of the antenna array. When electromagnetic waves are incident, the AMC structure has a high surface impedance, which can effectively reduce the backward radiation. It is also characterized by in-phase reflection, where the incident and reflected waves are superimposed to increase the forward gain.

In Fig. 8(a), the AMC reflection element is printed onto a 0.1 mm thick LCP dielectric substrate in the shape of a square patch. The metal ground plane completely covers the dielectric board’s rear surface. Figure 8(b) displays the AMC array’s corresponding circuit model. The AMC structure is equivalent to a LC parallel circuit. The gap between the patch and the ground will produce an equivalent capacitance C ₁, and there is also an equivalent capacitance C ₂ between two adjacent patches. The square patch on the upper layer generates an equivalent inductance L ₁.

The proposed AMC structure is an evolution of the traditional mushroom-shaped structure. The size of AMC structure can refer to the following formula [Reference Ashyap, Dahlan, Abidin, Abbasi, Kamarudin, Majid, Dahri, Jamaluddin and Alomainy24]:

(1)\begin{equation}{L_1} = {\mu _0}t\end{equation}

(2)\begin{equation}{f_0} = \frac{1}{{2\pi \sqrt {{L_1}{C_1}} }}\end{equation}

(3)\begin{equation}{C_1} = \frac{{{\varepsilon _0}\left( {1 + {\varepsilon _r}} \right)w}}{\pi }cos{h^{ - 1}}\left( {\frac{{w + g}}{g}} \right)\end{equation}

where ${L_1}$ and ${C_1}$ stand for the AMC structure’s inductance and capacitance, ${\varepsilon _0}$ is the dielectric constant of free space, ${\varepsilon _r}$ is the dielectric constant of the substrate, $w$ is the width of the AMC square patch, $g$ is the distance between the AMC patch, ${\mu _0}$ is the permeability of the dielectric substrate, $t$ is the dielectric substrate’s thickness, and ${f_0}$ is the operating frequency of the AMC structure. The optimized dimensions of the AMC element are as follows: w = 14.6 mm, g = 0.35 mm, t = 0.1 mm.

Figure 9 shows the design model and reflection phase characteristics of the AMC element, which has a reflection phase bandwidth of 60 MHz, covering 5.77–5.83 GHz. A model diagram of the proposed AMC array is given in Fig. 10. This AMC array consists of 9 × 3 units and the backside is completely covered by a metal ground.

Figure 9. Reflection phase of the AMC unit.

Figure 10. The structure of the AMC array. (a) Top view (b) Back view.

Antenna array loaded with AMC structure

The 9 × 3 AMC structure is loaded based on the antenna array, which is 11.8 mm away from the antenna array. During the simulation, the air-filled gap is between the array and AMC reflector. After loading the AMC reflector, the antenna’s total dimensions are 137.7 × 45.9 mm². The structure is shown in Fig. 11.

Figure 11. The AMC-loaded antenna array model.

The comparison of antenna S₁₁ parameters before and after loading AMC is displayed in Fig. 12. Between 5.57 and 6.11 GHz, the AMC-loaded antenna’s S₁₁ is below −10 dB and covers the WBAN operating frequency band of 5.725–5.875 GHz. The relative bandwidth reaches 9.2%. Figure 13 shows the radiation pattern of the antenna before and after loading the AMC reflector at 5.8 GHz. From the figure, we can conclude that the antenna’s backward radiation is significantly reduced after loading the AMC structure, and the forward gain increases. At 5.8 GHz, the antenna’s gain is 13.5 dB., which is 3.8 dB higher than that without the AMC reflector structure. Furthermore, the FBR improves by 26.04 dB.

Figure 12. S₁₁ parameters of the antenna.

Figure 13. Radiation patterns with and without AMC structure at 5.8 GHz. (a) E plane (b) H plane.

The surface currents at 5.8 GHz for the AMC unit and the array antenna loaded with the AMC structure are shown in Fig. 14. The maximum current density is observed at the floor gap section and at the front feeder. At the same scale, the surface current density of the 9 × 3 AMC structure placed behind the antenna array is significantly reduced and the back radiation is effectively reduced.

Figure 14. Surface current distribution. (a) AMC unit (b) Antenna array with AMC structure.

The effect of the distance H between the AMC and the antenna as well as the number of AMC units on the performance of the antenna is explored. Figure 15 gives the S₁₁ parameters of the antenna after loading the AMC structure at varying values of H. As H increases, the resonant frequency point moves to the left. The FBR and realized gain values of the antenna at different H are given in Table 1. When H is taken as 11.8 mm, the S₁₁ is from 5.58 to 6.1 GHz, the antenna realized gain is 13.5 dB, and the FBR is greatly improved to 27.48 dB, which can effectively reduce the radiation to the human body, so the optimum value of H is chosen as 11.8 mm.

Figure 15. The effect of H on S₁₁ parameters.

Table 1. The effect of H on FBR and realized gain values

To investigate the impact of varying ACM structures on the antenna, Fig. 16 provides the S₁₁ parameters and radiation patterns of the antenna loaded with various AMC structures. As the number of AMC units increases, the impedance matching of the antenna gets better and the S₁₁ value decreases gradually. The bandwidth and realized gain of the antenna with different sizes are given in Table 2. From the table it is clear that the bandwidth of the antenna is almost constant. With the enhancement of the AMC reflector plate size, the gain is improved. From Fig. 16(b) and (c), the EH-plane radiation patterns of the antenna are almost constant. Considering both size and performance, the AMC array size is finally selected 9 × 3.

Figure 16. Effect of loading different number of AMCs on antenna performance. (a) Effects on S₁₁ parameters (b) Effects on E-plane radiation (c) Effects on H-plane radiation.

Table 2. The performance of antennas with different numbers of AMC units

Figure 17 shows the S₁₁ values of the antenna loaded by the AMC and Perfect Electric Conductor (PEC)structure, and the PEC reflector is the same size as the AMC structure. From the figure, it can be seen when the distance between the antenna and the PEC structure is gradually increased, the resonant frequency point moves to the left. When the distance H is 11.8 mm, the S₁₁ of the antenna loaded with AMC structure is better than PEC structure. The FBR values of the antenna for different values of H are given in Table 3. When the H is 11.8 mm, the antenna loaded AMC structure is able to achieve higher FBR values than PEC structure. When the antenna is loaded by the PEC structure, the FBR value increases with the enhancement of H, but the space size is larger. In summary, the antenna loaded AMC structure can achieve a higher FBR value with a small size.

Figure 17. Simulated S₁₁ of the antenna supported by the AMC and PEC structure.

Table 3. The FBR values for different values of H

S₁₁ parameter

The proposed antenna was fabricated, as shown in Fig. 18. In a microwave anechoic chamber, the return loss and far-field radiation patterns of the antenna were measured.

Figure 18. Antenna physical diagram.

The return loss of both the standalone antenna array and the antenna array loaded with an AMC structure was tested using a vector network analyzer. Figure 19 shows the measurement results. The measured S₁₁ of the antenna array without AMC structure covers 5.22–5.88 GHz. After loading the AMC structure, the S₁₁ covers 5.62–6 GHz, achieving a relative bandwidth of 6.5%. Compared with the simulation data, the measured return loss and bandwidth exhibit some deviations, as well as the measured resonant frequency shifts to 5.8 GHz. These differences in measurement results may be due to machining errors and the test environment.

Figure 19. Comparison of measured and simulated data.

The antenna was placed on the back, chest, and legs of the human body, to test its performance. The measured S₁₁ is shown in Fig. 20, which is less than −10 dB in the required frequency band of 5.725–5.875 GHz, indicating good performance.

Figure 20. Measured S₁₁ parameters on different body parts.

Far-field radiation

The radiation pattern and gain of the antenna are tested in the far-field darkroom. Figure 21 displays the antennas’ 5.8 GHz simulated and measured radiation patterns. From the figure we can see that the test results are basically consistent with the simulation results. The antenna’s backward radiation is effectively reduced. After loading AMC, the simulated side lobe of the antenna is −17.6 dB, and the measured side lobe can only reach −13.1 dB due to the influence of test error.

Figure 21. Far-field radiation patterns at 5.8 GHz. (a) E plane (b) H plane.

In Fig. 22(a), the measured and simulated gain comparison of the antenna with and without AMC structure is given. After loading the AMC array, the gain is increased from 8.8 dB to 12.03 dB at 5.8 GHz, which is increased by 3.23 dB. The measured gain is reduced by about 1.5 dB compared with the simulation results. The variance in measurement and environmental uncertainty is the reason for the deviation of the results. In Fig. 22(b), the FBR value of the antenna array without AMC structure at 5.8 GHz is 1.44 dB. After loading AMC, the FBR becomes 27.48 dB at 5.8 GHz. In Fig. 23, the radiation efficiency plot of the antenna loaded with AMC structure is given. In the simulation results, the radiation efficiency is over 90%. The measured radiation efficiency of the antenna is higher than 80% within the desired frequency band 5.725–5.875 GHz, which provides good radiation results. The discrepancy between the over 90% simulated radiation efficiency and the smaller 80% measured radiation efficiency due to factors such as machining errors in processing the antenna, testing errors, and the environment surrounding the antenna during testing.

Figure 22. Comparison of antenna array performance with and without AMC. (a) Realized gain (b) FBR value.

Figure 23. Radiation efficiency of antenna loaded AMC structure.

SAR evaluation

As wearable device, the SAR of the antenna needs to be evaluated. The SAR value is a parameter to measure the amount of electromagnetic radiation energy absorbed by a substance. The US federal standards require that SAR values are less than 1.6 W/kg, while the European federal standard is not more than 2 W/kg.

Figure 24(a) shows a simulation model of SAR values based on 10 g of human tissue. A four-layer human tissue model is constructed in HFSS, which is skin, fat, muscle, and bone from the surface to the inside. The distance between the antenna array and human model is 3 mm in the simulation. Figure 24(b) shows the simulated S₁₁ of the antenna with and without the human body. Since the human model is also a medium when placed on a human body model for simulation, the S₁₁ becomes worse, but the operating frequency band is almost unaffected and can cover 5.55–6.15 GHz. The electromagnetic properties of the layers of the mannequin are shown in Table 4.

Figure 24. SAR value simulation. (a) Simulation model (b) S₁₁ parameters.

Table 4. Electromagnetic properties of various human tissues

Figure 25 shows the simulation results of the SAR value at 5.8 GHz. The SAR values with and without AMC structure are compared in the figure. According to the results, SAR field of antenna array on the human body with AMC structure is significantly reduced.

Figure 25. The changes of SAR value without and with loading AMC. (a) SAR value without AMC (b) SAR value with AMC.

Table 5 shows the effect of the size of the AMC array on SAR values. From the table it can be seen that the larger the size of the AMC array, the lower the SAR value. The SAR values with varying H are discussed in Table 6. It can be seen that the larger the distance H from the antenna to the human body, the smaller the SAR value.

Table 5. The effect of AMC array size on SAR values

Table 6. The effect of H on SAR values