Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Design of a transparent and flexible broadband omnidirectional antenna using characteristic mode analysis

Open Access Open Access

Abstract

An optically-transparent, mechanically flexible, monopole-radiation-characteristic, broadband omnidirectional antenna is proposed in this work. The radiator layer of the sandwich-structured antenna consists of a circular central patch with 16 circular satellite patches, and the characteristic mode analysis (CMA) was used to guide the detailed design process. We first analyzed the different modes of the 3×3 circular patch array, further reduced the size of the surrounding patch to increase the required omnidirectional mode bandwidth, further reduced the cross-polarization by adjusting the distance between the four corner patch and the central patch, and finally increased the number of surrounding patches to increase the bandwidth and gain. The entire analysis and optimization process uses the CMA. The antenna was excited by a single probe, with good omnidirectionality, low cross-polarization, and broad operating bandwidth in (5-12.7) GHz, achieving 87% relative bandwidth and peak gain of 4.1 dBi. The antenna has a radius of 27 mm and a thickness of 4.5 mm. A transparent and flexible antenna sample was fabricated on a flexible polydimethylsiloxane (PDMS) substrate using electrohydrodynamic (EHD) printing technology. The optical transmittance of the horseshoe-structured metal grids fabricated by EHD was investigated, as well as the equivalent surface resistance evaluation method. The theoretical optical transmittance of the horseshoe-structured single-layer silver grid on the PDMS substrate was 81.2%, versus 80.8% measured value, with an equivalent surface resistance of roughly 5 $\Omega /\square$. The proposed antenna was bent on a 50 mm-radius cylinder surface and retained excellent omnidirectionality.

© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

1. Introduction

Compact broadband antennas with monopole radiation characteristics have become a research hotspot in recent years, mainly because of their vertical polarization (VP) omnidirectional radiation characteristics, which can provide complete signal coverage in 360 degrees [15]. The conventional monopole whip antenna cannot completely meet the application requirements due to its high-profile and narrow-band properties [6]. At present, many works have been proposed to improve the antenna bandwidth [711]. But low-profile monopole antennas usually require complex feed matching networks and struggle to maintain acceptable omnidirectionality with low cross-polarization over the operation band [12]. Considering the diversity of application environments, the design and fabrication of broadband omnidirectional antennas with small size, low profile, simple feed, and low cross-polarization are very attractive.

At the same time, due to its potential applications in fields including vehicle networking, biomedicine, and satellite communications, optically-transparent and mechanically-flexible antennas have attracted increasing attention in recent years [1317]. Polydimethylsiloxane (PDMS), as an excellent transparent dielectric material, has been widely used in flexible antennas due to its stable dielectric properties, high optical transparency, good flexibility and excellent biocompatibility [1820]. The transparent conductor, usually used as the antenna’s radiator as well as the ground layers, is another key factor in the transparent and flexible antenna design. Metal grid films have a huge potential in the antenna design, because of its adjustable optical transparency, excellent electric conductivity, and good ductility [14,2123]. In general, metal grid conductive films offer lower ohmic loss while preserving good optical transparency than conventional conductive films, resulting in higher radiation gain, but the fabrication is not easy [24].

This work describes the design and fabrication of a transparent, flexible, omnidirectional broadband antenna with monopole-like radiation properties. The proposed antenna was fabricated on a PDMS substrate, and the radiator and ground layers were made of a horseshoe-shaped optically-transparent silver grid that was directly printed on the top and bottom surfaces of PDMS using the electrohydrodynamic (EHD) printing technology [25,26]. The EHD printing technology is one of the most promising developments in the high-precision printing-electronics area, and has been successfully applied in the fabrication of high-precision metal grids, sensors, transparent microwave devices, and so on [2729]. In this work, we innovatively used EHD printing technology to produce transparent flexible antenna with high-precision horseshoe shaped grids. Both simulation and experiment results verified the antenna’s monopole radiation characteristics in a flat state as well as on a cylindrical surface with a radius of 50 mm, and also demonstrate the excellent flexibility of the proposed antenna.

2. Design of the proposed antenna

The schematic diagram of the proposed sandwich-structure antenna is shown in Fig. 1. The top radiator layer contains a central circular patch with 16 satellite circular patches, and Table 1 shows the dimension details after optimization. The dielectric substrate of the antenna is PDMS with a thickness of 4.5 mm, a dielectric constant of 2.75, and a loss tangent of 0.02. The radiator pattern and the ground layer are printed in the form of a horseshoe-shaped silver grid at the top and bottom of the substrate. This feeding design effectively bypasses possible connection problem between the center patch and the probe during antenna bending. To improve its adaptability to harsh environments, a protective layer of 25 $\mathrm{\mu}$m PDMS is applied on the top and bottom surfaces, respectively.

 figure: Fig. 1.

Fig. 1. Schematic design of the proposed sandwich-structure antenna. For clarity, the 25 $\mathrm{\mu}$m protective PDMS layers are not included.

Download Full Size | PDF

Tables Icon

Table 1. Structure Parameters of the Proposed Sandwich-Structure Antenna

3. Operating mechanisms and radiator layer design

3.1 Theory of characteristic modes

The characteristic mode (CM) theory was first proposed in 1971 for conductor analysis [30] and has since been extensively utilized to antenna design [3136]. In this section, we will briefly introduce some important parameters of CM. The linear superposition of the currents from all modes can be used to express the total current $J$ in an ideal conductor:

$$J=\sum_{n=1}^{N}{{{\alpha }_{n}}{{J}_{n}}}$$

The contribution of the ${n^{th}}$ mode ${{J}_{n}}$ to the overall current is described by the mode weight coefficient (MWC) ${{\alpha }_{n}}$ :

$${{\alpha }_{n}}=\frac{1}{(1+j{{\lambda }_{n}})\int{{{J}_{n}}\centerdot {{E}_{i}}dS}}$$
where ${{\lambda }_{n}}$ is the eigen value of mode ${{J}_{n}}$, ${{E}_{i}}$ is the incident electric field, and $S$ is the conductor surface.

In the absence of an excitation input, the mode significance (MS), also known as the inherent potential contribution of each mode to the structure radiation.

$$MS=\frac{1}{|1+j{{\lambda }_{n}}|}$$

To determine the electric feeding, the mode excitation coefficient (MEC) $v_{n}^{i}$, which is defined as:

$$v_{n}^{i}=\langle {{J}_{n}},{{E}_{i}}\rangle =\mathop{{\int\!\!\!\!\!\int}\mkern-21mu \bigcirc}\nolimits_S {{J}_{n}}\centerdot {{E}_{i}}dS$$

In the antenna design using the characteristic mode analysis (CMA), the half-power bandwidth is defined as:

$$BW=\left| \frac{{{f}_{h}}-{{f}_{l}}}{{{f}_{c}}} \right|$$
where ${{f}_{h}}$, ${{f}_{l}}$ and ${{f}_{c}}$ represent the highest frequency, lowest frequency and center frequency in the MS$\ge$0.707 frequency band, respectively.

3.2 CMA of the original radiator layer

The original radiator layer was first studied to initiate the antenna design process, as shown in Fig. 2. The radius r of the circular patch is determined according to the following equation [37].

$$r=\frac{{{k}_{01}}c}{2\pi {{f}_{01}}\sqrt{{{\varepsilon }_{r}}}}$$
where ${{k}_{01}}$=2.405 is the first root of the ${{0}^{th}}$ order Bessel function, $r$ is the radius of the circular patch, ${{f}_{01}}$ is the corresponding resonant frequency.

 figure: Fig. 2.

Fig. 2. Original radiator layer design: the circular patches are evenly arranged forming a square array.

Download Full Size | PDF

The radiator layer consists of 3$\times$3 circular patches with radius $r={{r}_{c}}=13 mm$, gap $g$=1 mm, and dielectric size $p$=85 mm, without considering the feeding structure. The commercial simulation software CST MWS 2021 utilizes a multilayer solver to provide the simulation results. The boundary condition settings are as follows: "open boundary" in the x, y, and +z direction, and "infinite floor" in the -z direction. Unless otherwise noted, all simulation frequencies are between 2 and 15 GHz, with the mode number ordered at 5.5 GHz.

The MSs of the first 15 modes are shown in Fig. 3. ${{J}_{n({{r}_{c}}=m)}}$ is used to denote the nth mode at $r_c$=m mm to distinguish the mode number of various structures. The fundamental mode is shifted to lower frequencies due to the addition of outer patches. ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{6({{r}_{c}}=13)}}$ resonate around 5.5 GHz and ${{J}_{7({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$ resonate around 3.5 GHz. Figure 4 shows the mode current distribution and radiation patterns of ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$ at the resonant frequency. Degenerate modes include ${{J}_{3({{r}_{c}}=13)}}$ and ${{J}_{4({{r}_{c}}=13)}}$, ${{J}_{7({{r}_{c}}=13)}}$ and ${{J}_{8({{r}_{c}}=13)}}$, ${{J}_{9({{r}_{c}}=13)}}$ and ${{J}_{10({{r}_{c}}=13)}}$. Among these 15 modes, ${{J}_{12({{r}_{c}}=13)}}$ and ${{J}_{13({{r}_{c}}=13)}}$ are expected to achieve good omnidirectional radiation through reasonable structural optimization and feed configuration. Compared to ${{J}_{12({{r}_{c}}=13)}}$, ${{J}_{13({{r}_{c}}=13)}}$ is easier to be excited by simple feeding. Therefore, ${{J}_{13({{r}_{c}}=13)}}$ is the ideal mode for the desired VP omnidirectional radiation with an MS bandwidth of 11% and a frequency range of (3.6-4.5) GHz. Figure 5 shows the mode current distribution and radiation patterns of ${{J}_{13({{r}_{c}}=13)}}$ in (3.5-5) GHz. The current is mainly distributed on the outer circular patch, and the antenna is difficult to excite. It can be seen from above that the original design features narrow MS bandwidth and difficult excitation, so does not match the broadband design requirements.

 figure: Fig. 3.

Fig. 3. MSs of the original radiator layer from ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$.

Download Full Size | PDF

 figure: Fig. 4.

Fig. 4. Diagrams of the mode current distribution and radiation patterns from ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$.

Download Full Size | PDF

 figure: Fig. 5.

Fig. 5. Diagrams of the mode current distribution and radiation patterns of ${{J}_{13({{r}_{c}}=13)}}$ at different frequencies.

Download Full Size | PDF

To excite omnidirectional radiation in a broadband, a feed structure that can efficiently excite a specific mode with a broadband MS is required. Two commonly used feed sources are the magnetic current source and the electric current source, which correspond to two basic feed structures, slot and probe. Complicated feeds can degrade the mechanical strength of flexible transparent antennas and complicate the fabrication process. In this work, a probe is used as the single-point feed at the center of the antenna, which necessitates a broadband MS and a high concentration of modal currents in the central patch.

3.3 CMA of the outer patch size

The resonant frequency and mode current distribution can be modified by varying the size of the outer patches. For optimization of the outer patch sizes, the following analysis is performed. Due to the decrease of the outer patch size, the resonant frequency is shifted to higher frequency, as shown in Fig. 6(a). The mode ordering is also changed, and ${{J}_{10({{r}_{c}}=8)}}$ and ${{J}_{5({{r}_{c}}=3)}}$ correspond to ${{J}_{13({{r}_{c}}=13)}}$. Furthermore, the MS bandwidth corresponding to the three modes is enhanced significantly, and by modifying the size of the outer patch, higher MS may be obtained in the desired broadband frequency range.

 figure: Fig. 6.

Fig. 6. The MS spectra at different ${{r}_{c}}$. (b) Spectra of inner-to-outer current ratio K at different ${{r}_{c}}$.

Download Full Size | PDF

The current distribution is shown by the inner-to-outer patch current ratio $K$ [38]:

$$K=\frac{{{A}_{inner}}}{{{A}_{outer}}}$$
where ${{A}_{inner}}$ and ${{A}_{outer}}$ are the maximum values of the current amplitude in the inner and outer patches, respectively. The higher the value of $K$ is, the more current is concentrated in the center patch. $K$>1 indicates that the current is mainly concentrated in the center patch. When the outer patch size decreases, the current distribution shifts from being mostly dispersed in the outer patches to being mostly concentrated in the center patch, as shown in Fig. 6(b).

3.4 CMA of the distance from the four corner patches to the center patch

After decreasing the substrate length $p$ to 54 mm and setting $r_c$=3 mm, the current distribution and radiation patterns of ${{J}_{5({{r}_{c}}=3)}}$ (corresponding to ${{J}_{13({{r}_{c}}=13)}}$ of the original structure) at different frequencies are analyzed as below.

The current is mainly concentrated in the center patch at low frequencies, and the mode has excellent omnidirectionality, as shown in Fig. 7, where ${{J}_{5({{r}_{c}}=3)}}$ at 5 GHz and 8 GHz is TM$_{01}$-like and TM$_{02}$-like, respectively. The cross-polarization of ${{J}_{5({{r}_{c}}=3)}}$ at 5 GHz and 8 GHz is <−30 dB, as shown in Fig. 8. The current is mostly along the radial direction and the tangential component is very tiny, guaranteeing low cross-polarization. When the frequency exceeds 11 GHz, the current on the outer patches becomes stronger, and anisotropic current appears on the center patch, causing the radiation pattern to deform and the omnidirectionality to degrade. At the same time, the tangential component of the current increases, increasing the antenna’s cross-polarization.

 figure: Fig. 7.

Fig. 7. Mode current distribution profiles and radiation patterns of ${{J}_{5({{r}_{c}}=3)}}$ at 5, 8, 11, 14 GHz.

Download Full Size | PDF

 figure: Fig. 8.

Fig. 8. Normalized horizontal $(\theta =30{}^\circ )$ radiation pattern of ${{J}_{5({{r}_{c}}=3)}}$ at 5, 8, 11, 14 GHz.

Download Full Size | PDF

Due to their separation from the center patch, the four corner patches and other nearby patches have an anisotropic current distribution, which is the primary contributor to the distortion of the radiation pattern. The anisotropic current distribution further causes the tangential component of the center patch current to increase, resulting in high cross-polarization. To reduce high frequency cross-polarization, the four corner circular patches are positioned nearer the center patch until their separation from it is equal to that of the patches around it. Define $J_{n({{r}_{c}}=m)}^{'}$ as the current distribution of mode n when the distance from all surrounding patches, including the four corner patches, to the center patch is equal and $r_c$=m mm, as in Fig. 9 and 10. When the frequency is high (11 GHz or 14 GHz in the figures), the current direction of $J_{5({{r}_{c}}=3)}^{'}$ in the outer patch remains unchanged, and the tangential current of the center patch disappears, considerably improving the antenna’s omnidirectionality and cross-polarization compared to the original design.

 figure: Fig. 9.

Fig. 9. Mode current distribution profiles and radiation patterns of $J_{5({{r}_{c}}=3)}^{'}$ at 5, 8, 11, 14 GHz.

Download Full Size | PDF

 figure: Fig. 10.

Fig. 10. Normalized horizontal $(\theta =30{}^\circ )$ radiation pattern of ${{J}_{5({{r}_{c}}=3)}}$ at 11, 14 GHz.

Download Full Size | PDF

3.5 CMA of the improved antenna

To optimize the omnidirectionality and the cross-polarization, the substrate shape is changed from rectangular to circular, with ${{r}_{5}}$=27 mm, ${{r}_{1}}$=0.635 mm, $d$=0.35 mm, and the other dimensions stay unchanged. Define $J_{n({{r}_{c}}=m)}^{''}$ as the current distribution of mode n when ${{r}_{c}}$ =m mm under center probe excitation. Figure 11(a) and (b) depict the MS and MWC of the first 6 modes. It can be seen that mode 1 dominates below 6.9 GHz and mode 6 dominates above 6.9 GHz. The surface current distribution, $J_{1({{r}_{c}}=3)}^{''}$ and $J_{6({{r}_{c}}=3)}^{''}$, are depicted in Fig. 12(a) and (b). $J_{1({{r}_{c}}=3)}^{''}$ behaves like TM${_{01}}$; $J_{6({{r}_{c}}=3)}^{''}$ behaves as TM$_{01}$-like below 8 GHz, TM${_{02}}$-like around 11 GHz, and TM${_{03}}$-like around 14 GHz. The center probe feeding can excite the omnidirectional modes in a broadband.

 figure: Fig. 11.

Fig. 11. (a)MS and (b)MWC spectra of the first 6 modes for the circular-shape substrate design, with center probe excitation.

Download Full Size | PDF

 figure: Fig. 12.

Fig. 12. (a) Current distribution profiles of mode 1 at different frequencies. (b) Current distribution profiles of mode 6 at different frequencies.

Download Full Size | PDF

3.6 Effect of the quantity of outer circular patches

Full-wave simulations were performed to further illustrate the influence of the outer circular patch quantity on the relative bandwidth and peak gain. Broadband impedance matching can be achieved by adjusting the radius of the outer circular patches ${{r}_{c}}={{r}_{4}}-{{r}_{3}}$ , the distance between the center patch and the surrounding patches $\left ( {{r}_{3}}-{{r}_{2}} \right )$, and the feed gap d. The optimization results as follows can lead a broadband impedance matching: $~{{r}_{4}}-{{r}_{3}}=3$ mm, ${{r}_{3}}-{{r}_{2}}=0.5$ mm, $d$=0.35 mm, and the surface resistance of both the radiator and the ground is 5 $\Omega /\square$. Figure 13(a) shows that for the number of outer circles $n$=8, 12, and 16, the relative bandwidth is 85.6%, 87%, and 88%, respectively, and the peak gain at 10 GHz is 5.15 dBi, 5.43 dBi, and 5.6dBi, respectively. Enhanced bandwidth and gain within the broadband are achieved by increasing the number of outer circular patches, which does not impair the antenna’s omnidirectionality due to the geometric symmetry. Figure 13(b) shows the radiation efficiency curve of the designed antenna in the (5-14) GHz when $n$=16.

 figure: Fig. 13.

Fig. 13. (a) |S11| and gain of antennas with different number of outer patches. (b) Radiation efficiency curve of the designed antenna in the (5-14) GHz when $n$=16.

Download Full Size | PDF

4. Fabrication of the transparent flexible antenna

4.1 Antenna fabrication

As shown in Fig. 14, the antenna sample was fabricated on 4.5 mm-thick PDMS with a dielectric constant of 2.75 and a loss tangent of 0.02. The PDMS film was laser cut into a circle with an outer diameter ${{r}_{5}}=\text { }27$ mm and an inner diameter ${{r}_{1}}=\text { }0.635$ mm. The PDMS surface is activated using a spray-type AP plasma activator (CRF, Guangdong, China). The plasma activator’s power is set to 1000 W. The nozzle is 20 mm high above the PDMS surface and moves along the bow at a speed of 50 mm/s. Ensure the nozzle movement range is greater than the PDMS sample at the same time. To improve the PDMS’s hydrophobicity, repeat the previous steps three times. EHD printing technology can not only print large area of fine structures, but also can directly print patterns on flexible substrates and complex surfaces without making expensive etching templates. It shows advantageous application prospect in the field of flexible and transparent antenna preparation with low manufacturing cost. So the EHD technology was used to print a horseshoe-shaped structure silver grid with optimal dimensions on the PDMS surface, as illustrated in Fig. 15. A protective 25 $\mathrm{\mu}$m PDMS layer was placed on top and bottom of the main-part PDMS, respectively. The ground of the SMA connector was fixed to the ground layer by the nickel-carbon conductive silicone rubber, and the SMA-to-PDMS-substrate connection was further reinforced with the non-conductive 703 silicone rubber for extra mechanical strength. There was no electrical contact between the probe and the radiator layer.

 figure: Fig. 14.

Fig. 14. (a) Photograph of the optimized antenna sample (the logo of the Nanjing University of Science and Technology printed on a piece of paper can be clearly seen through the sample). (b) Photograph of the antenna sample under test in the microwave anechoic chamber.

Download Full Size | PDF

 figure: Fig. 15.

Fig. 15. (a) Illustration of the horseshoe-shaped grid structure. (b) Single-cycle dissection of the horseshoe-shaped grid structure. (c) Illustration of the dimension parameters of the horseshoe structure. (d) Micrograph of the horseshoe-shaped silver grid fabricated by EHD printing (scale bar: 200 $\mathrm{\mu}$m).

Download Full Size | PDF

4.2 Optical transmittance of the horseshoe-shaped silver grid

A horseshoe-shaped silver grid structure was utilized instead of the conventional cross-straight-line grid structure to enhance the antenna’s bending performance, as shown in Fig. 15(a). As shown in Fig. 15(b), the horseshoe structure can be divided into individual periodic cells, and the total area $S$ enclosed by the four inverted S-shaped structures, the linewidth area $S1$, and the theoretical duty cycle or effective area ratio T can be calculated as below:

$$S\approx {{(4\cdot rb-(rb-ra))}^{2}}+4\cdot 2\pi \cdot ra\cdot \frac{360{}^\circ{-}{{\theta }_{1}}}{360{}^\circ }\cdot (rb-ra) (8)$$
$$S1\approx 8\cdot 2\pi \cdot ra\cdot \frac{360{}^\circ{-}{{\theta }_{1}}}{360{}^\circ }\cdot (rb-ra)$$
$$T\approx \frac{{{(4\cdot rb-(rb-ra))}^{2}}-4\cdot 2\pi \cdot ra\cdot \frac{360{}^\circ{-}{{\theta }_{1}}}{360{}^\circ }\cdot (rb-ra)}{{{(4\cdot rb-(rb-ra))}^{2}}+4\cdot 2\pi \cdot ra\cdot \frac{360{}^\circ{-}{{\theta }_{1}}}{360{}^\circ }\cdot (rb-ra)}$$

Figure 15(c) shows a horseshoe structure with a line width of 20 $\mathrm{\mu}$m, ${{r}_{a}}=240$ $\mathrm{\mu}$m, ${{r}_{b}}=260$ $\mathrm{\mu}$m, and $\theta =170{}^\circ$. According to (10), the theoretical optical transmittance of the single-layer horseshoe grid without the PDMS substrate is 88.47%, and that with the PDMS substrate is 81.2%. Figure 15(d) shows the micrograph of the EHD-printed horseshoe-shaped structure. The optical transmittance of the PDMS substrate itself was 91.8% (haze=0.26), measured by the WGT-S Transmittance/haze Tester (INESA, Shanghai, China), while the optical transmittance of the sample with the single-layer horseshoe-shaped silver grid printed on the PDMS substrate was 80.8% (haze=1.26), which is very close to the theoretical transmittance of 81.2%.

4.3 Equivalent surface resistance of the horseshoe-shaped silver grid

In the simulation, for easy and efficient optimization, the horseshoe-shaped silver grid was represented by a simple resistive sheet, whose equivalent surface resistance is the same as the silver grid. In fabrication, it is important to assess and control the equivalent surface resistance of the horseshoe-shaped silver grid. Figure 16 shows the equivalent electric circuit for a M$\times$N grid structure. The resistance between two arbitrary points $\alpha$, $\beta$ in the resistive network can be calculated from the following equation [39]:

$${{R}_{\alpha \beta }}=\sum_{i=2}^{\mathcal{N}}{\frac{1}{{{l}_{i}}}}{{\left| {{\psi }_{i\alpha }}-{{\psi }_{i\beta }} \right|}^{2}}$$

 figure: Fig. 16.

Fig. 16. Equivalent electric circuit for a M$\times$N grid structure.

Download Full Size | PDF

Here ${{l}_{i}}$ stands for the nonzero eigenvalues of the Laplace operator, and the corresponding eigenvectors are ${{\psi }_{i}}=({{\psi }_{11}},{{\psi }_{12}},{\ldots } ,{{\psi }_{1\mathcal {N}}})$, $i=2,3,{\ldots } ,\mathcal {N}$. The number of nodes of the two-dimensional resistive network is (M+1) $\times$ (N+1), and the nodes can be numbered by coordinates m, n, where $0\le m\le M,\text { }0\le n\le N$. The eigenvalues and eigenvectors of the Laplace operator can be calculated as below:

$${{\lambda }_{(m,n)}}=2{{r}^{{-}1}}(1-\cos {{\theta }_{m}})+2{{s}^{{-}1}}(1-\cos {{\phi }_{n}})$$
$$\psi _{(m,n);(x,y)}^{free}=\psi _{mx}^{(M+1)}\psi _{ny}^{(N+1)}$$
where
$${{\theta }_{m}}=\frac{m\pi }{M+1}$$
$${{\phi }_{n}}=\frac{n\pi }{N+1}$$

The resistance ${{R}_{\text {free}}}$ between two arbitrary nodes ${{{r}}_{1}}=\left ( {{x}_{1}},{{y}_{1}} \right )$ and ${{{r}}_{2}}=\left ( {{x}_{2}},{{y}_{2}} \right )$ is

$$\begin{aligned} &R_{\{{{({M + 1})} \times {({N + 1})}}\}}^{free}\left( {r_{1},r_{2}} \right)\\ &= {\sum_{m = 0}^{M}{{\sum_{n = 0}^{N}~}_{{({m,n})} \neq {(0,0)}}\frac{\left| {\psi_{{({m,n})};{({x_{1},y_{1}})}}^{free} - \psi_{{({m,n})};{({x_{2},y_{2}})}}^{free}} \right|^{2}}{\lambda_{({m,n})}}}}\\ &= \frac{r}{N + 1}\left| {x_{1} - x_{2}} \right| + \frac{s}{M + 1}\left| {y_{1} - y_{2}} \right| + \frac{2}{(M + 1) \times (N + 1)}\\ &\times {\sum_{m = 1}^{M}{\sum_{n = 1}^{N}\frac{\left\lbrack {\cos\left( {x_{1} + \frac{1}{2}} \right)}\theta_{m}{\cos\left( {y_{1} + \frac{1}{2}} \right)}\phi_{n} - {\cos\left( {x_{2} + \frac{1}{2}} \right)}\theta_{m}{\cos\left( {y_{2} + \frac{1}{2}} \right)}\phi_{n} \right\rbrack^{2}}{r^{- 1}\left( {1 - cos\theta_{m}} \right) + s^{- 1}\left( {1 - cos\phi_{n}} \right)}}} \end{aligned}$$

Here $r$, $s$ are the resistance values for a single period along the two main directions. Because of symmetry, $r$ equals to $s$ when $M+1=N+1$, ${{x}_{1}}={{y}_{1}},{{x}_{2}}={{y}_{2}}$. The resistance $r$ between the two nodes ${{r}_{1}}$ and ${{r}_{2}}$ can be directly measured by the multimeter. The transverse resistance of the entire resistive network is generated by the parallel connection of (M+1) resistors with resistance value M${\times }r$, assuming that the current is in the transverse direction and ignoring the longitudinal resistance. The equivalent surface resistance of the grid structure can be obtained from the following equation (16):

$$R=\frac{M}{M+1}r$$

We printed numerous groups of horseshoe-shaped silver grid with 10$\times$10 periods on the PDMS substrate, then measured and calculated the surface resistance using the above method. The equivalent surface resistance of the horseshoe-shaped grid was verified to be around 5 $\Omega /\square$.

5. Simulation and experiment results

The transparent and flexible sample fabricated in section 4 was measured to verify the analysis and design of the proposed antenna. As shown in Fig. 17, the |S11| of the proposed antenna was measured using an Agilent N5244A network analyzer. Take |S11|<−10 dB as the effective radiation standard, the simulated radiation band was (5.1-13) GHz, while the measured radiation band was (5-12.7) GHz. The measurement results matched well with the simulation, with a relative bandwidth of 87%. The measured gain in the operating frequency band ranged from −2.5 dBi to 4.1 dBi, versus (0.13-5.6) dBi from simulation. The possible reasons of why the measured gain is lower than the simulation include that the resistance between the SMA connector and the ground layer was not included in the simulation, and the actual surface resistance at high frequency may be greater than that used in the simulation. In addition, the fabrication errors would lead to imperfect uniformity of the grid conductivity, which could result in worse symmetry of radiator layer and broader lobe width.

 figure: Fig. 17.

Fig. 17. |S11| and gain of the proposed antenna.

Download Full Size | PDF

The normalized radiation patterns of the vertical $\left ( \varphi =\text { }0{}^\circ \right )$ and horizontal $\left ( \theta =\text { }30{}^\circ \right )$ planes of the proposed antenna at 6,10, and 12 GHz are presented in Fig. 18, respectively, to demonstrate the antenna’s omnidirectionality and low cross-polarization characteristics. The cross-polarization radiation levels in both vertical and horizontal planes were <−10 dB, and roundness deviation in the horizontal plane was <6 dB. The symmetry of the designed structure was not perfect due to fabrication errors, which is the primary reason why the cross-polarization radiation level and roundness are worse than the simulation values. The overall simulation and measurement results, however, are generally consistent, and the antenna exhibits a monopole-like radiation mode in a broadband.

 figure: Fig. 18.

Fig. 18. Simulated and measured normalized radiation patterns of the proposed antenna at (a) 6 GHz (b) 10 GHz (c) 12 GHz.

Download Full Size | PDF

To test the antenna performance under physical deformation, the sample was bent and attached on the side of a foam cylinder (cross-sectional radius 50 mm), as illustrated in Fig. 19(a). Figure 19(b) compares the normalized horizontal plane radiation pattern at 10 GHz for the flat and bent states. The results reveal that after bending, the proposed antenna still has good omnidirectionality.

 figure: Fig. 19.

Fig. 19. (a) Schematic diagram of the antenna sample in flat and bent states; (b) The normalized radiation patterns of the antenna sample in the horizontal plane $\left ( \theta =\text { }30{}^\circ \right )$ in flat and bent states.

Download Full Size | PDF

Table 2 compares the proposed antenna’s performance to that of other broadband omnidirectional antennas. This work features a small size, a broad relative bandwidth, a simple feeding mechanism, and good radiation performance. What is more, it is the only flexible and transparent device in the table.

Tables Icon

Table 2. Comparison of Some Reported Omnidirectional Antennas

6. Conclusions

A transparent and flexible sandwich-structure omnidirectional antenna was proposed in this study, and a sample was fabricated by the EHD technology. The antenna was thoroughly investigated and optimized using CMA. A special horseshoe-shaped silver grid structure was designed and analyzed to form the radiator and ground layers, so as to achieve low surface resistance (5 $\Omega /\square$), good optical transparency (>80%) as well as considerable mechanical flexibility (50 mm bending radius) at the same time. The operating bandwidth of the proposed antenna was (5-12.7) GHz, i.e., 87% relative bandwidth, and the peak gain was 4.1 dBi. The proposed antenna maintains low cross-polarization and good omnidirectionality in the whole operating frequency band, and can maintain good omnidirectionality under bending.

Funding

Fundamental Research Funds for the Central Universities (30917012202, 30920021128); National Natural Science Foundation of China (61871225, 61901219).

Acknowledgements

This work was supported by the Fundamental Research Funds for the Central Universities under Grant Number 30917012202, the Fundamental Research Funds for the Central Universities under Grant Number 30920021128, the National Natural Science Foundation of China under Grant Number 61901219, and the National Natural Science Foundation of China under Grant Number 61871225.

Disclosures

The authors declare no conflicts of interest.

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

References

1. T. Aboufoul, C. Parini, X. Chen, and A. Alomainy, “Pattern-reconfigurable planar circular ultra-wideband monopole antenna,” IEEE Trans. Antennas Propag. 61(10), 4973–4980 (2013). [CrossRef]  

2. A. Dastranj and B. Abbasi-Arand, “High-performance 45° slant-polarized omnidirectional antenna for 266-GHz UWB applications,” IEEE Trans. Antennas Propag. 64(2), 815–820 (2016). [CrossRef]  

3. K. Ghaemi and N. Behdad, “A low-profile, vertically polarized ultrawideband antenna with monopole-like radiation characteristics,” IEEE Trans. Antennas Propag. 63(8), 3699–3705 (2015). [CrossRef]  

4. S. Nikolaou and M. A. B. Abbasi, “Design and development of a compact UWB monopole antenna with easily-controllable return loss,” IEEE Trans. Antennas Propag. 65(4), 2063–2067 (2017). [CrossRef]  

5. J. de Dieu Ntawangaheza, L. Sun, S. Wang, Y. Li, Z. Zheng, D. Biao, and G. Rushingabigwi, “A single-layer low-profile broadband metasurface-inspired antenna with monopole-like radiation characteristics,” IEEE Trans. Antennas Propag. 70(6), 4818–4823 (2022). [CrossRef]  

6. M. Li and N. Behdad, “A compact, capacitively fed UWB antenna with monopole-like radiation characteristics,” IEEE Trans. Antennas Propag. 65(3), 1026–1037 (2017). [CrossRef]  

7. M. Alibakhshikenari, B. S. Virdee, V. Vadala, M. Dalarsson, M. E. d. C. Gomez, A. G. Alharbi, S. N. Burokur, S. Aissa, I. Dayoub, F. Falcone, and E. Limiti, “Broadband 3-d shared aperture high isolation nine-element antenna array for on-demand millimeter-wave 5g applications,” Optik 267, 169708 (2022). [CrossRef]  

8. M. Alibakhshikenari, B. S. Virdee, A. A. Althuwayb, L. Azpilicueta, N. O. Parchin, C. H. See, R. A. Abd-Alhameed, F. Falcone, I. Huynen, T. A. Denidni, and E. Limiti, “Bandwidth and gain enhancement of composite right left handed metamaterial transmission line planar antenna employing a non foster impedance matching circuit board,” Sci. Rep. 11(1), 7472 (2021). [CrossRef]  

9. M. Alibakhshikenari, B. S. Virdee, P. Shukla, Y. Wang, L. Azpilicueta, M. Naser-Moghadasi, C. H. See, I. Elfergani, C. Zebiri, R. A. Abd-Alhameed, I. Huynen, J. Rodriguez, T. A. Denidni, F. Falcone, and E. Limiti, “Impedance bandwidth improvement of a planar antenna based on metamaterial-inspired t-matching network,” IEEE Access 9, 67916–67927 (2021). [CrossRef]  

10. M. Alibakhshi-Kenari, M. Naser-Moghadasi, R. A. Sadeghzadeh, B. S. Virdee, and E. Limiti, “Miniature crlh-based ultra wideband antenna with gain enhancement for wireless communication applications,” ICT Express 2(2), 75–79 (2016). [CrossRef]  

11. M. Alibakhshi-Kenari, M. Naser-Moghadasi, R. A. Sadeghzadeh, and B. S. Virdee, “Hexa-band planar antenna with asymmetric fork-shaped radiators for multiband and broadband communication applications,” IET Microwaves Antennas & Propag. 10(5), 471–478 (2016). [CrossRef]  

12. K. S. Ryu and A. A. Kishk, “Uwb dielectric resonator antenna having consistent omnidirectional pattern and low cross-polarization characteristics,” IEEE Trans. Antennas Propag. 59(4), 1403–1408 (2011). [CrossRef]  

13. A. S. M. Alqadami, K. S. Bialkowski, A. T. Mobashsher, and A. M. Abbosh, “Wearable electromagnetic head imaging system using flexible wideband antenna array based on polymer technology for brain stroke diagnosis,” IEEE Trans. Biomed. Circuits Syst. 13(1), 124–134 (2019). [CrossRef]  

14. T. Jang, C. Zhang, H. Youn, J. Zhou, and L. J. Guo, “Semitransparent and flexible mechanically reconfigurable electrically small antennas based on tortuous metallic micromesh,” IEEE Trans. Antennas Propag. 65(1), 150–158 (2017). [CrossRef]  

15. A. S. M. Sayem, R. B. V. B. Simorangkir, K. P. Esselle, and R. M. Hashmi, “Development of robust transparent conformal antennas based on conductive mesh-polymer composite for unobtrusive wearable applications,” IEEE Trans. Antennas Propag. 67(12), 7216–7224 (2019). [CrossRef]  

16. A. S. M. Sayem, R. B. V. B. Simorangkir, K. P. Esselle, R. M. Hashmi, and H. Liu, “A method to develop flexible robust optically transparent unidirectional antennas utilizing pure water, pdms, and transparent conductive mesh,” IEEE Trans. Antennas Propag. 68(10), 6943–6952 (2020). [CrossRef]  

17. Z. H. Jiang, Z. Cui, T. W. Yue, Y. Zhu, and D. H. Werner, “Compact, highly efficient, and fully flexible circularly polarized antenna enabled by silver nanowires for wireless body-area networks,” IEEE Trans. Biomed. Circuits Syst. 11(4), 920–932 (2017). [CrossRef]  

18. R. B. V. B. Simorangkir, Y. Yang, K. P. Esselle, and B. A. Zeb, “A method to realize robust flexible electronically tunable antennas using polymer-embedded conductive fabric,” IEEE Trans. Antennas Propag. 66(1), 50–58 (2018). [CrossRef]  

19. Y. Liu, Q. Wang, Y. Jia, and P. Zhu, “A frequency- and polarization-reconfigurable slot antenna using liquid metal,” IEEE Trans. Antennas Propag. 68(11), 7630–7635 (2020). [CrossRef]  

20. Z. Hu, C. Chen, Z. Zhou, and Y. Li, “An epsilon-near-zero-inspired pdms substrate antenna with deformation-insensitive operating frequency,” Antennas Wirel. Propag. Lett. 19(9), 1591–1595 (2020). [CrossRef]  

21. P. D. Tung and C. W. Jung, “Optically transparent wideband dipole and patch external antennas using metal mesh for uhd tv applications,” IEEE Trans. Antennas Propag. 68(3), 1907–1917 (2020). [CrossRef]  

22. S. Hong, Y. Kim, and C. W. Jung, “Transparent microstrip patch antennas with multilayer and metal-mesh films,” Antennas Wirel. Propag. Lett. 16, 772–775 (2017). [CrossRef]  

23. S. H. Kang and C. W. Jung, “Transparent patch antenna using metal mesh,” IEEE Trans. Antennas Propag. 66(4), 2095–2100 (2018). [CrossRef]  

24. G. Clasen and R. Langley, “Meshed patch antennas,” IEEE Trans. Antennas Propag. 52(6), 1412–1416 (2004). [CrossRef]  

25. W. Gu, P. E. Heil, H. Choi, and K. Kim, “Generation of stable multi-jets by flow-limited field-injection electrostatic spraying and their control via i-v characteristics,” J. Phys. D: Appl. Phys. 43(49), 492001 (2010). [CrossRef]  

26. W. Gu, P. E. Heil, H. Choi, and K. Kimb, “Comprehensive model for fine coulomb fission of liquid droplets charged to Rayleigh limit,” Appl. Phys. Lett. 91(6), 064104 (2007). [CrossRef]  

27. Y. Wu, Y. Deng, J. Wang, Z. Zong, X. Chen, and W. Gu, “Thz broadband absorber fabricated by EHD printing technology with high error tolerance,” IEEE Trans. Terahertz Sci. Technol. 9(6), 637–642 (2019). [CrossRef]  

28. Y. Wu, C. Fu, S. Qian, Z. Zong, X. Wu, Y. Yue, and W. Gu, “Flexible and transparent W-band absorber fabricated by EHD printing technology,” Antennas Wirel. Propag. Lett. 19(8), 1345–1349 (2020). [CrossRef]  

29. X. Zhu, Z. Qian, X. Chen, L. Liu, C. Sheng, and W. Gu, “Electrohydrodynamics-printed silver nanoparticle flexible pressure sensors with improved gauge factor,” IEEE Sens. J. 21(5), 5836–5844 (2021). [CrossRef]  

30. R. J. Garbacz and R. H. Turpin, “A generalized expansion for radiated and scattered fields,” IEEE Trans. Antennas Propag. 19(3), 348–358 (1971). [CrossRef]  

31. C. Fu, C. Feng, W. Chu, Y. Yu, X. Zhu, and W. Gu, “Design of a broadband high-gain omnidirectional antenna with low cross polarization based on characteristic mode theory,” Antennas Wirel. Propag. Lett. 21(9), 1747–1751 (2022). [CrossRef]  

32. T. Li and Z. N. Chen, “Wideband sidelobe-level reduced ka-band metasurface antenna array fed by substrate-integrated gap waveguide using characteristic mode analysis,” IEEE Trans. Antennas Propag. 68(3), 1356–1365 (2020). [CrossRef]  

33. T. Li and Z. N. Chen, “Shared-surface dual-band antenna for 5g applications,” IEEE Trans. Antennas Propag. 68(2), 1128–1133 (2020). [CrossRef]  

34. J. Wang, W. Wang, A. Liu, M. Guo, and Z. Wei, “Broadband metamaterial-based dual-polarized patch antenna with high isolation and low cross polarization,” IEEE Trans. Antennas Propag. 69(11), 7941–7946 (2021). [CrossRef]  

35. J. Zeng, X. Liang, L. He, F. Guan, F. H. Lin, and J. Zi, “Single-fed triple-mode wideband circularly polarized microstrip antennas using characteristic mode analysis,” IEEE Trans. Antennas Propag. 70(2), 846–855 (2022). [CrossRef]  

36. Q. Zhang, R. Ma, W. Su, and Y. Gao, “Design of a multimode uwb antenna using characteristic mode analysis,” IEEE Trans. Antennas Propag. 66(7), 3712–3717 (2018). [CrossRef]  

37. S. Liu, W. Wu, and D.-G. Fang, “Wideband monopole-like radiation pattern circular patch antenna with high gain and low cross-polarization,” IEEE Trans. Antennas Propag. 64(5), 2042–2045 (2016). [CrossRef]  

38. S. Liu, D. Yang, Y. Chen, K. Sun, X. Zhang, and Y. Xiang, “Design of single-layer broadband omnidirectional metasurface antenna under single mode resonance,” IEEE Trans. Antennas Propag. 69(10), 6947–6952 (2021). [CrossRef]  

39. F. Y. Wu, “Theory of resistor networks: the two-point resistance,” J. Phys. A: Math. Gen. 37(26), 6653–6673 (2004). [CrossRef]  

40. B. Mohamadzade, R. B. V. B. Simorangkir, R. M. Hashmi, and A. Lalbakhsh, “A conformal ultrawideband antenna with monopole-like radiation patterns,” IEEE Trans. Antennas Propag. 68(8), 6383–6388 (2020). [CrossRef]  

41. X. Yan, Y. Liu, and S.-X. Gong, “Design of a wideband omnidirectional antenna with characteristic mode analysis,” Antennas Wirel. Propag. Lett. 17(6), 993–997 (2018). [CrossRef]  

42. R. Lian, T. Shih, Y. Yin, and N. Behdad, “A high-isolation, ultra-wideband simultaneous transmit and receive antenna with monopole-like radiation characteristics,” IEEE Trans. Antennas Propag. 66(2), 1002–1007 (2018). [CrossRef]  

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (19)

Fig. 1.
Fig. 1. Schematic design of the proposed sandwich-structure antenna. For clarity, the 25 $\mathrm{\mu}$m protective PDMS layers are not included.
Fig. 2.
Fig. 2. Original radiator layer design: the circular patches are evenly arranged forming a square array.
Fig. 3.
Fig. 3. MSs of the original radiator layer from ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$.
Fig. 4.
Fig. 4. Diagrams of the mode current distribution and radiation patterns from ${{J}_{1({{r}_{c}}=13)}}$ to ${{J}_{15({{r}_{c}}=13)}}$.
Fig. 5.
Fig. 5. Diagrams of the mode current distribution and radiation patterns of ${{J}_{13({{r}_{c}}=13)}}$ at different frequencies.
Fig. 6.
Fig. 6. The MS spectra at different ${{r}_{c}}$. (b) Spectra of inner-to-outer current ratio K at different ${{r}_{c}}$.
Fig. 7.
Fig. 7. Mode current distribution profiles and radiation patterns of ${{J}_{5({{r}_{c}}=3)}}$ at 5, 8, 11, 14 GHz.
Fig. 8.
Fig. 8. Normalized horizontal $(\theta =30{}^\circ )$ radiation pattern of ${{J}_{5({{r}_{c}}=3)}}$ at 5, 8, 11, 14 GHz.
Fig. 9.
Fig. 9. Mode current distribution profiles and radiation patterns of $J_{5({{r}_{c}}=3)}^{'}$ at 5, 8, 11, 14 GHz.
Fig. 10.
Fig. 10. Normalized horizontal $(\theta =30{}^\circ )$ radiation pattern of ${{J}_{5({{r}_{c}}=3)}}$ at 11, 14 GHz.
Fig. 11.
Fig. 11. (a)MS and (b)MWC spectra of the first 6 modes for the circular-shape substrate design, with center probe excitation.
Fig. 12.
Fig. 12. (a) Current distribution profiles of mode 1 at different frequencies. (b) Current distribution profiles of mode 6 at different frequencies.
Fig. 13.
Fig. 13. (a) |S11| and gain of antennas with different number of outer patches. (b) Radiation efficiency curve of the designed antenna in the (5-14) GHz when $n$=16.
Fig. 14.
Fig. 14. (a) Photograph of the optimized antenna sample (the logo of the Nanjing University of Science and Technology printed on a piece of paper can be clearly seen through the sample). (b) Photograph of the antenna sample under test in the microwave anechoic chamber.
Fig. 15.
Fig. 15. (a) Illustration of the horseshoe-shaped grid structure. (b) Single-cycle dissection of the horseshoe-shaped grid structure. (c) Illustration of the dimension parameters of the horseshoe structure. (d) Micrograph of the horseshoe-shaped silver grid fabricated by EHD printing (scale bar: 200 $\mathrm{\mu}$m).
Fig. 16.
Fig. 16. Equivalent electric circuit for a M$\times$N grid structure.
Fig. 17.
Fig. 17. |S11| and gain of the proposed antenna.
Fig. 18.
Fig. 18. Simulated and measured normalized radiation patterns of the proposed antenna at (a) 6 GHz (b) 10 GHz (c) 12 GHz.
Fig. 19.
Fig. 19. (a) Schematic diagram of the antenna sample in flat and bent states; (b) The normalized radiation patterns of the antenna sample in the horizontal plane $\left ( \theta =\text { }30{}^\circ \right )$ in flat and bent states.

Tables (2)

Tables Icon

Table 1. Structure Parameters of the Proposed Sandwich-Structure Antenna

Tables Icon

Table 2. Comparison of Some Reported Omnidirectional Antennas

Equations (17)

Equations on this page are rendered with MathJax. Learn more.

J = n = 1 N α n J n
α n = 1 ( 1 + j λ n ) J n E i d S
M S = 1 | 1 + j λ n |
v n i = J n , E i = S J n E i d S
B W = | f h f l f c |
r = k 01 c 2 π f 01 ε r
K = A i n n e r A o u t e r
S ( 4 r b ( r b r a ) ) 2 + 4 2 π r a 360 θ 1 360 ( r b r a ) ( 8 )
S 1 8 2 π r a 360 θ 1 360 ( r b r a )
T ( 4 r b ( r b r a ) ) 2 4 2 π r a 360 θ 1 360 ( r b r a ) ( 4 r b ( r b r a ) ) 2 + 4 2 π r a 360 θ 1 360 ( r b r a )
R α β = i = 2 N 1 l i | ψ i α ψ i β | 2
λ ( m , n ) = 2 r 1 ( 1 cos θ m ) + 2 s 1 ( 1 cos ϕ n )
ψ ( m , n ) ; ( x , y ) f r e e = ψ m x ( M + 1 ) ψ n y ( N + 1 )
θ m = m π M + 1
ϕ n = n π N + 1
R { ( M + 1 ) × ( N + 1 ) } f r e e ( r 1 , r 2 ) = m = 0 M n = 0 N   ( m , n ) ( 0 , 0 ) | ψ ( m , n ) ; ( x 1 , y 1 ) f r e e ψ ( m , n ) ; ( x 2 , y 2 ) f r e e | 2 λ ( m , n ) = r N + 1 | x 1 x 2 | + s M + 1 | y 1 y 2 | + 2 ( M + 1 ) × ( N + 1 ) × m = 1 M n = 1 N [ cos ( x 1 + 1 2 ) θ m cos ( y 1 + 1 2 ) ϕ n cos ( x 2 + 1 2 ) θ m cos ( y 2 + 1 2 ) ϕ n ] 2 r 1 ( 1 c o s θ m ) + s 1 ( 1 c o s ϕ n )
R = M M + 1 r
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.