Antipodal Linearly Tapered Slot Antenna with Quasi-Hemispherical Pattern Using Subwavelength Elements

Abstract

Antennas with quasi-hemispherical radiation patterns are preferred in many wide−area wireless communication systems which require the signals to uniformly cover a wide two−dimensional region. In this work, a simple but effective beamwidth broadening technique based on an antipodal linearly tapered slot antenna (ALTSA) is first proposed and then experimentally verified. Compared with most of the reported designs, the proposed antenna can significantly widen beamwidth and achieve a quasi-hemispherical radiation pattern without increasing the overall size and structural complexity. Only two rows of subwavelength metallic elements (eight elements in total) are simply and skillfully printed at specified positions on the dielectric substrate (relative permittivity ε_r = 2.94 and thickness h = 1.5 mm) of a general ALTSA whose peak gain is 11.7 dBi, approximately 200% half-power beamwidth (HPBW) enlargement can be obtained in all cut-planes containing the end-fire direction at the central frequency of 15 GHz, and the HPBW extensions in different cut-planes have good consistency. Thus, a quasi-hemispherical beam pattern can be acquired. Thanks to the simplicity of this method, the antenna size and structural complexity do not increase, resulting in the characteristics of easy fabrication and integration, being lightweight, and high reliability. This proposed method provides a good choice for wide−beam antenna design and will have a positive effect on the potential applications of wide-area wireless communication systems.

1. Introduction

Nowadays, wide-beam antennas have a range of applications requiring wide coverage [1,2,3,4,5,6,7], such as satellite systems [8], radar detection [9,10], and 5G wireless communication [11,12,13]. Although wide beamwidth can be obtained by reducing antenna gain, the directivity of the antenna will also be reduced and cause a large loss of radiation energy. To solve this contradiction between high directivity and wide beamwidth, beam broadening technology is worth studying. In most of the reported studies, beamwidth widening is only aimed at either one or both of the E- and H-planes, while few pieces of research focus on all cut-planes to achieve a quasi-hemispherical radiation pattern, especially for end-fire antennas. Meanwhile, to achieve the goal of beam broadening, most of the reported studies have paid the cost of overall size and structural complexity increase. All of these insufficiencies limit the applications of wide-beam antennas.

In recent years, a large number of techniques have been investigated to broaden beamwidth. In [14], a binary coding technique was proposed to determine the complex metal patterns of a patch antenna. In this way, the beamwidth in the H−plane for port 1 was effectively broadened. However, it also significantly increased the design complexity. The authors of [15] employed four separated microstrip patches to load on a dielectric substrate as the radiators. By adjusting the distance between these patches, the beamwidth in the E-plane could be widened. Unfortunately, the antenna suffered from design complexity, and beam broadening in all cut-planes could not be achieved. In [16], a new aperture-coupled cylindrical dielectric resonator antenna was designed. By means of exciting the basic and high-order modes in the dielectric resonator, a widened beam could be realized in the E- and H-planes. Nonetheless, the overall size of the antenna was enlarged due to the dielectric resonator loading, and the stacking of multi-layered cylindrical disks also increased the manufacturing difficulty. Another dielectric resonator antenna was also proposed in [17]. Owing to the loading of the engraved groove and comb-like metal wall, the HPBW in the E- and H- planes could be widened, but at the cost of the size and design complexity. In [18], a magnetoelectric dipole could achieve a stable wide beam in the H-plane; regrettably, however, the HPBW extension was not large enough and the structure was complex. The authors of [19] combined a U-shaped reflector with L-shaped metal arms as a magnetic dipole and added a parasitic patch to the L-shaped metal arms as an electric dipole. Thereby, the impedance bandwidth and the antenna HPBW were broadened. Nevertheless, a larger antenna size was needed and fabrication at high frequencies was not easy. In [20], to achieve a broad beamwidth, the authors combined a parasitic strip with a rectangular microstrip magnetic dipole antenna (MMDA). The MMDA was equivalent to a magnetic dipole toward the y-axis direction, and the parasitic strip worked as an electric dipole toward the x-axis direction. The beamwidth in the yoz cut-plane could be widened from about 70° to 180° due to the superposition of multiple complementary sources. However, the beamwidth was basically widened in a single cut-plane. In [21], a subarray, including a main radiant patch and two parasitic patches, was created to widen the beamwidth in the E-plane. By adjusting the spacing between the patches, the amplitude and phase of the coupling energy from the main patch to the parasitic patches could be controlled and the superposed radiation patterns could be affected. Nevertheless, this also only widened beamwidth in a single cut-plane. In [22], the mushroom structure and a pair of equal-amplitude and out-of-phase Huygens sources were introduced into the dielectric resonator antenna. The beamwidth was widened from 90° to 194°. However, the beamwidth could only be widened within one cut-plane.

The antipodal linearly tapered slot antenna (ALTSA), developed from the Vivaldi aerial [23], has numerous excellent qualities, such as good radiation orientation, low cost, being lightweight, easy fabrication, and easy integration, etc. The combination of the ALTSA and wide-beam antenna has the potential to achieve broader applications. However, research on this subject is scarce. In [24,25], artificial material units were loaded on the surface of an antipodal tapered slot antenna (ATSA) to improve directivity. Nevertheless, these methods further narrowed the beamwidth and required the printing of a large number of units, which increased the overall size. In [26], a smart configuration was proposed. Two antipodal tapered patches were designed as a fan-like shape for widening the beam, and good results were realized. However, the HPBW amplification was limited, and the design was also complex.

It can be observed that none of the above designs focus on widening the beams uniformly in all cut-planes without sacrificing antenna size and complexity, especially for end-fire antennas. This limits the applications of wide-beam antennas.

In this paper, a beam-widening design based on an ALTSA is first proposed. By loading two rows of subwavelength metallic strips on the blank area between the two tapers of the basic ALTSA, the HPBW can be widened by about 200% in entire cut-planes that include the boresight at the central frequency of 15 GHz, and the difference between each HPBW amplification is small. Thus, a quasi-hemispherical pattern can be acquired. In this way, there is almost no increase in overall size and structural complexity. Along with the above characteristics, the proposed design also has the advantages of being low cost, lightweight, and having strong reliability, which makes it a good candidate for a wide-beam antenna.

2. Antenna Configuration and Performance Analysis

2.1. Antenna Configuration

The configurations of the proposed wide-beam antenna are presented in Figure 1, which comprises a basic ALTSA and a pair of subwavelength elements. For this basic ALTSA, as shown in Figure 1a, two symmetrical metal tapers with slots are printed on the top and bottom side of the dielectric plate (relative permittivity ε_r = 2.94 and thickness h = 1.5 mm) as radiators. The microstrip line serves as a feed for both tapers, and the substrate-integrated waveguide (SIW) structure is used to optimize the impedance matching between them. The slots in the metal tapers are employed to reduce the antenna size and optimize the antenna pattern, where their width, depth, and spacing are indicated.

A pencil beam toward the boresight is radiated by the basic ALTSA. To widen the beamwidth, two oblique rows of subwavelength metal units, as shown in Figure 1b, are loaded on the dielectric substrate of the basic ALTSA. The length of the subwavelength metal unit is about 0.3λ₀, where λ₀ is the free-space wavelength at the central frequency of 15 GHz. Each row has four units, and the subwavelength structures are designed as simple I-shapes. Since there are only eight subwavelength elements and they are etched on the surfaces of the dielectric substrate, the overall size and structural complexity of the antenna are hardly increased.

After the combination of the basic ALTSA and the subwavelength metal units, the proposed antenna in our article is formed, as shown in Figure 1c, and the wide−beam effect can be realized. Figure 1d shows the exploded view. From these two Figures, it can be seen that the subwavelength elements are attached to the blank area between the two tapers near the end of the antenna. It is notable that the two groups of subwavelength units are not integrated into the same plane, but are printed separately on the top and bottom sides of the dielectric substrate together with one of the tapers. Additionally, the closer to the end of the antenna, the closer the subwavelength units are to the metal tapers, but they do not touch. Figure 1e shows the local geometries of the proposed wide−beam antenna, and the optimized parameter values are shown in

Table 1. Optimized geometrical parameters (unit: mm).

Parameter	W_floor	W_exc	D_hole	d_slot	W_slot
Value	12.5	3.4	1	0.5	3.5
Parameter	L_slot	D	W	L	t
Value	4.5	5.5	2.9	3.43	0.06

.

2.2. Simulated Results and Analysis

The simulations and optimizations of the proposed antenna were carried out in Ansoft High-Frequency Structure Simulator (HFSS). The proposed antenna can effectively broaden the beamwidth in the frequency band of 14.5–15.5 GHz. Figure 2 and Figure 3 show the simulation results of |S₁₁| and the gain patterns of the proposed antenna. Meanwhile, the corresponding simulation results of the basic ALTSA are also presented for comparison. As shown in Figure 2, the |S₁₁| of these two antennas were both less than −10 dB in the band of 14.5–15.5 GHz, indicating a good impedance matching. Considering the characteristic of the narrow band, the far-field radiation performances of the wide-beam antenna were demonstrated and analyzed only at the center frequency f_c = 15 GHz. Figure 3 shows the gain results of these two antennas in each cut-plane, including the boresight (+z-axis direction). We present patterns in cut-plane increments of 15 degrees (phi) to illustrate the performance. Phi represents the azimuth angle, phi = 0° represents the xoz-plane, phi = 90° represents the yoz-plane, and the coordinate system is given in Figure 1. For observing the amplification of HPBW in each cut-plane more directly, the relationship between the amplification of HPBW and phi is shown in Figure 4.

It can be seen from Figure 3 and Figure 4 that the HPBW in each cut-plane of the basic ALTSA was 36–45° with a peak gain of 11.7 dBi, while the HPBW of the proposed antenna was 65–97°. The red and blue horizontal lines represent a gain drop of 3 dB for the proposed antenna and the basic antenna, respectively. The HPBW in each cut-plane was widened, and all widening ranges were about 200%, showing good consistency. The beamwidths of the basic ALTSA were relatively small due to their high gain, which also limits the absolute values of the broadened beamwidth. Proper selection of a basic antenna with a lower gain is beneficial to obtain a larger absolute value of beamwidth.

It should be pointed out that this research aims to demonstrate a new beam−broadening idea based on an ALTSA. Depending on the practical beamwidth requirements, other basic ALTSAs which have different gain and beamwidth can be flexibly selected. Moreover, the basic antenna and subwavelength elements in this paper were designed with simple structures and ordinary performances. The achievements of beam broadening can be further enhanced if basic ALTSA and subwavelength units with more complicated structures and better performance are chosen.

Considering that the number of subwavelength elements loaded has a great impact on the performances of the antenna, parameter analysis was carried out. Figure 5 shows the performances of antennas with different numbers of subwavelength elements in each row. As shown in Figure 5a, when there were three or four units in each row, the |S₁₁| of the antenna were almost lower than −10 dB in the whole frequency band, indicating good impedance matching. Additionally, as shown in Figure 5b, when there were two to four units in each row, the amplification of HPBW in all cut-planes was similar, but the amplification of HPBW with four units in each row was the largest. Hence, four units in each row were selected.

To qualitatively analyze the reason why HPBW was broadened, the E-field amplitude and phase results of the basic ALTSA and the proposed wide−beam antenna are given, as shown in

Table 2. E−field amplitude and phase distributions of the proposed and basic antennas.

	Proposed Antenna	Basic Antenna
Amplitude
Phase (Phi = 0)
Phase (Phi = 45)
Phase (Phi = 90)
Phase (Phi = 135)

. At the same time, considering that the performances in each cut-plane were relatively close, only the corresponding phase diagrams in the phi = 0°, 45°, 90°, and 135° cut-planes at 15 GHz are presented here. As can be seen from Table 2, before loading subwavelength units, the E-field energy distribution between the two tapers of basic ALTSAs was relatively uniform, and the beam energy was mainly shot toward the boresight. However, after adding subwavelength units, the E−field energy distribution in this area significantly changed, and the energy was mainly concentrated near the subwavelength units, deviating from the boresight and transmitting to both sides. Similarly, the subwavelength elements change the equivalent dielectric constant of the relevant region, thus affecting the transmission phase of the electromagnetic wave. Additionally, the shape of the wavefront then changed, making the beam radiate to both sides. These are beneficial to broadening the beamwidth. The current distribution of the proposed antenna at 15 GHz is also given in Figure 6. From Figure 6, we can observe that the subwavelength units had a strong current response, which indicates that the subwavelength units play an important role in widening beamwidth.

2.3. Performance Comparison

The performance of the proposed antenna in this work were compared with other wide−beam antennas in publications, as shown in

Table 3. Comparison between wide-beam antennas.

Ref.	Type	Radiation Direction	Aim at All Cut-Planes	# HPBW	* Increased Size	* Increased Complexity
[2]	Dielectric resonator antenna	Broadside	No	189%	Yes	Yes
[14]	Patches antenna	Broadside	No	~200%	No	Yes
[17]	Dielectric resonator antenna	Broadside	No	~350%	Yes	Yes
[19]	Dipole antenna	Broadside	No	~230%	Yes	Yes
[26]	Tapered slot antenna	End-fire	No	<200%	No	Yes
[27]	Microstrip antenna	Broadside	No	~230%	Yes	Yes
[28]	Dielectric resonator antenna	Broadside	No	~330%	Yes	Yes
[29]	Microstrip antenna	Broadside	No	241%	No	Yes
This work	ALTSA	End-fire	Yes	237%	No	No

#: Maximum amplification of the HPBW in the given cut-planes mentioned in the literature. *: Increased size or complexity compared to basic antennas, and little addition treated as no increase. ~: Not mentioned directly but estimated according to the figures in the reference.

. It can be seen that there are few instances of research into broadening HPBW in all cut-planes, especially with end-fire characteristics. Moreover, most of the reported articles could not achieve beam broadening without sacrificing antenna size and complexity. For example, in [14,17,19], some interesting designs were presented that could effectively achieve wide beams, but all experienced the disadvantages of increased size, complexity, or both. In this work, based on an ALTSA with an end-fire pattern, two rows of quasi-intersecting subwavelength elements were loaded, realizing a wide beam in all cut−planes without increasing the antenna size and complexity, and the extensions in different cut-planes were similar. As mentioned above, the proposed antenna meets the demands of wide beam, low cost, and easy fabrication in many applications.

3. Fabrication and Measurement

To prove the properties of the proposed wide-beam antenna, a prototypical antenna was fabricated using the printed-circuit-board (PCB) process, and measured in the microwave anechoic chamber, as shown in Figure 7. The top and bottom view of the fabricated wide-beam antenna are presented in Figure 7a,b, respectively, and Figure 7c shows the measurement scene. In this section, we present the measured reflection coefficient in Figure 8. It can be observed that the |S₁₁| were less than −10 dB in the whole frequency band. Additionally, we also display the measured gain results in different cut-planes at the central frequency and compare them with the simulated ones. Considering that Section 2 has given the simulated gain results, where phi was chosen from 0° to 180° at an interval of 15°, the HPBW amplifications in different cut-planes were relatively consistent. Therefore, the gain consequences only in the phi = 0°, 45°, 90°, and 135° cut-planes are demonstrated here, as shown in Figure 9. It can be seen that the measured and simulated results had a general consistency, with slight deviations caused by the dielectric loss and fabrication tolerances. Taking the phi = 0° cut-plane results as an example, the measured HPBW was about 69°, covering a range from −32° to 37°, with a peak gain of 6.1 dBi. Meanwhile, the simulated HPBW in this cut-plane was approximately 78°, overriding a scale from −38° to 40°, with a peak gain of 6.7 dBi. The difference between the measured and simulated HPBW was about 9°. On the whole, the measured results are in good agreement with the simulated results.

4. Conclusions

In most of the reported studies, beam-widening could only be realized in one or both of the E- and H-planes, and came at the cost of increased antenna size and complexity. These obviously limit the applications of wide-beam antennas. To address these problems, in this paper, a novel and simple method to widen the antenna beamwidth was put forward. By simply printing two quasi-intersecting rows of subwavelength cells on the dielectric substrate of a basic ALTSA, the beamwidth can be widened in all cut-planes that include the boresight direction with a tiny increase in antenna size and complexity. Additionally, the widening amplitudes at the center frequency are stable at around 200%, which produces a quasi-hemispherical pattern. Moreover, the main goal of this paper is to show the feasibility of this beam-broadening method. It is completely possible to use the basic ALSTA and subwavelength units with more complex structures and excellent performances to obtain better results. The designed antenna also has the characteristics of being low-cost and lightweight, easy to fabricate, and easy to integrate, making it a good candidate for satellite communication antennas.