I. Introduction
In recent years, global climate change driven by rapid global warming has accelerated manifold. Consequently, damage caused by typhoons, strong winds, high waves, and rip currents has increased [1, 2]. Coastal erosion caused by high waves, which degrade coastal land and result in loss of sand, has been consistently observed. Coastal erosion alters coastal ecosystems and damages the surrounding infrastructure, leading to collapsed buildings. It also greatly affects the disaster risks and livelihoods of residents in coastal areas. For example, it has a detrimental impact on economic activities, such as tourism and fishing, in coastal areas.
To mitigate these effects, extensive research is being conducted worldwide to accurately observe and analyze wave heights—the primary cause of coastal erosion—and to minimize wave-induced damage [3–6]. The primary system used for this purpose is an underwater ultrasonic sensor system, which was developed in 1977 [7]. Ultrasonic sensor systems are primarily used in harsh environments, such as the seafloor, to measure distances and detect objects or terrain, including wave movements. However, when installed on the seafloor, expert divers are required for both the installation and retrieval of such systems, posing an elevated risk of accidents. Additionally, installation and data retrieval costs are high. Moreover, sea surface bubbles can compromise the accuracy of wave observation measurements.
To address these drawbacks, X-band navigation radar systems, which utilize wave radar cross-section (RCS) information, were developed for wave observation [8–10]. Compared to an ultrasonic sensor system, an X-band navigation radar system is installed on land and uses electromagnetic waves, making it safer than an underwater ultrasonic sensor system. In addition to wave observations, it offers the advantages of being multipurpose and suitable for coastal surveillance and navigation [11–15]. However, X-band navigation radars offer limited capabilities in terms of wave observation owing to their low frequency with regard to the RCS of waves, which makes it challenging to achieve a wide bandwidth. This results in difficulties with real-time large-scale signal processing and reduces the accuracy of wave height and vertical movement observations.
To perform real-time signal processing of big data generated by nonlinearly moving waves, it is essential to develop a radar system that can operate in the higher Ku-band frequency range rather than the X-band frequency range that is currently used [16–18]. Moreover, to achieve high performance in radar systems, the performance of the antenna, which is a key component of radar, is crucial. Among the various antenna performance characteristics—gain, beam width, bandwidth, efficiency, pattern, and reflection coefficient—the gain is particularly important because it affects the resolution and observational range of radar systems. Therefore, designing a high-gain antenna is integral.
Table 1 presents a comparison of the antenna parameters of the conventional X-band marine radars used for maritime applications and those of the proposed Ku-band radar. As indicated in Table 1, the antenna developed for X-band radar operates between 9.3 GHz and 9.4 GHz, with a bandwidth of 100 MHz. In contrast, the proposed Ku-band radar antenna has a wide bandwidth of 500 MHz, meaning that it offers five times the frequency bandwidth of X-band radars. This wide bandwidth is essential for the real-time signal processing of big data related to waves. In terms of gain, which affects the resolution of the radar, the Ku-band antenna is superior to the X-band antenna by approximately 7.8 dBi. Moreover, since the X-band radar antenna is primarily used for navigation, it is characterized by horizontal polarization. In contrast, the Ku-band radar for wave monitoring is designed with vertical polarization, which is critical for observing high wave heights. In addition, the Ku-band radar antenna is designed with a wider half-power beam width (HPBW) in the vertical plane and a narrower HPBW in the horizontal plane than the X-band radar antenna.
Based on this comparison, the proposed Ku-band radar antenna can be considered more suitable for real-time observations and signal processing of big data on wave height. In particular, its broadband and high-gain design plays a crucial role. In this context, slotted waveguide array antennas are excellent candidates for achieving high gain with low loss in high-frequency bands [19–21]. This is because a large number of resonant slots results in a large slot aperture, thereby providing better performance [22, 23].
In this paper, we investigate a method for improving antenna gain that positions a grid cavity array on top of the radiating waveguide [24] to suppress sidelobes and achieve high gain. This method involves designing a conventional grid cavity as a structure integrated with the radiating waveguide. However, impedance variations were observed owing to design errors that arose during the process of separately fabricating and then assembling the grid cavity and radiating waveguide. A consistent discrepancy was also identified between the measured and simulated reflection coefficients.
In this paper, these shortcomings are addressed by optimally designing the grid cavity and radiating waveguide by accounting for the errors that may occur during antenna fabrication. Additionally, the mutual impedance effects are analyzed. The proposed approach improves the reflection coefficient characteristics by combining the grid cavity and radiating waveguide, thereby enhancing the agreement between the simulation and measurement results. The proposed antenna is designed using Ansys HFSS—a 3D simulation program based on the finite element method. The center frequency obtained is 17.5 GHz, and the target frequency band is 500 MHz, ranging from 17.25 GHz to 17.75 GHz. The lengths of the broad and narrow walls of the designed waveguide are 10.79 mm and 5.4 mm, respectively, with the guided wavelength, λg, being 28.21 mm.
A detailed comparison of the design results for the proposed 6 × 26 slotted waveguide array antenna and those for the antenna in [25] is presented in Section II, while the design and measurement results for the grid cavity are presented in Sections III and IV, respectively.
II. Slotted Waveguide Array Antenna Design
Fig. 1 illustrates the design structure of the 6 × 26 slotted waveguide array antenna. Fig. 1(a) depicts the top view of the antenna, highlighting the positions of the resonant slots and feeding points, the direction of the electric field radiation pattern, and the connection holes for combining the grid cavity. Fig. 1(b) presents a side view of the antenna, showing the design for feeding a plane wave into the waveguide using an SMA connector as the input port. Fig. 1(c) presents the parameters of the resonant slots and the spacing between the slots, with the slot spacing fixed at λg/2 to maintain the same phase. Since slot length (Sl) affects resonant frequency, iterative calculations were performed to determine the optimal slot length and width (Sw). Considering the Sw, the Sl was found to be optimal in the range of λg/4–3λg/8. Furthermore, to obtain the optimum resonant slot parameters by conducting iterating calculations, the optimal width was maintained in the range of λg/25–λg/15. Given that the slot offset (So) impacts the real part of the impedance, it also has a significant effect on the reflection coefficient. Overall, the optimal values of Sl, Sw, and So were obtained through iterative calculations. Fig. 1(d) depicts the isometric view of the radiating section, where the waveguide’s broad wall (WG_a), narrow wall (WG_b), and waveguide thickness (t) are indicated. The parameters calculated for the proposed antenna are listed in Table 2.
Fig. 2 presents a comparison of the reflection coefficients of the antenna in [25] and the redesigned antenna. Since the former antenna lacked sufficient space to accommodate the grid cavity proposed in this paper, it had to be redesigned. The blue dotted and black dashed lines in Fig. 2 represent the simulated and measured results of the antenna in [25], respectively, while the red solid line denotes the simulated results of the proposed antenna. Although the design parameters of the two antennas differ slightly, both were designed to resonate at the center frequency and meet the target frequency bandwidth.
Table 2 indicates that to secure space for integrating the grid cavity in the radiating section, the slot offset was redesigned to be larger than that of the antenna in [25]. Consequently, the slot offset was increased from 0.66 mm to 1.15 mm. Owing to this change, the parameters for the slot length, width, waveguide broad wall, waveguide narrow wall, and thickness had to be adjusted. Furthermore, as shown in Fig. 2, the reflection coefficients of the proposed antenna match well with those of the antenna in [25].
Fig. 3 compares a snapshot of the simulated surface current distribution of the waveguide antennas at the center frequency of 17.5 GHz. Fig. 3(a) and 3(b) illustrate the surface current distribution of the antenna in [25] and that of the proposed antenna, respectively. Fig. 3(b) demonstrates a more concentrated surface current distribution compared to Fig. 3(a), because the slot width of the proposed antenna was designed to be larger than that of the antenna in [25]. Consequently, the aperture of the proposed antenna was larger. Based on this result, the gain of the proposed antenna, as shown in Fig. 3(b), was expected to be slightly higher at 17.5 GHz.
Fig. 4(a) and 4(b) compare the simulated electric field radiation patterns Eθ and Eφ of the antennas at 17.5 GHz, respectively. The blue dotted and red solid lines represent the simulation results of the antenna in [25] and those of the proposed antenna, respectively. The beam patterns of both the reference and proposed antennas show similar results.
Fig. 5 presents a comparison of the simulated peak gain across the target frequency range of the antennas. The blue dotted and black dashed lines represent the simulation and measurement results of the antenna in [25], respectively, while the solid red line indicates the simulated results of the proposed antenna. It is evident that the gains of both antennas exceed 28 dBi. At the center frequency, the gain of the proposed antenna is approximately 1.31 dBi—higher than that of the reference antenna.
III. Grid Cavity Design
1. Grid Cavity Parameter Study
Fig. 6 presents the structure of the antenna proposed in Section II, with the grid cavity combined on top of the radiating section, along with a detailed explanation of the grid cavity. In this structure, the role of the grid cavity is to suppress sidelobes and improve gain. The diffracted fields that occur near the four corners of a radiating antenna greatly contribute to the generation of sidelobes [26, 27]. Therefore, a grid cavity was designed to suppress these sidelobes. The key parameters of the grid cavity design include its height (Gh), horizontal width (Ghw), and vertical width (Gvw). These parameters were optimized by performing iterative calculations.
Fig. 7(a) compares the reflection coefficients obtained by varying the height of the grid cavity. Notably, the height of the grid cavity greatly affects the imaginary components of impedance and phase change, causing observable frequency shifts and changes in reflection coefficients. Theoretically, at a height of 7 mm, as represented by the thin black dotted line close to λg/4 in Fig. 7(a), no phase change difference was observed, and optimal impedance matching was expected. However, owing to the small width of the grid cavity, an optimal design required a height shorter than λg/4. Consequently, the best results were achieved for a height of 5 mm, as indicated by the solid red line.
Fig. 7(b) compares the reflection coefficients obtained for variations in the horizontal width of the grid cavity. Iterative calculations were performed to determine the optimal horizontal width, with the grid cavity height and vertical width fixed at 5.0 mm and 2.5 mm, respectively. It is evident that the horizontal width significantly affects the imaginary part of the impedance, resulting in sensitive variations in frequency shifts. Resonance occurred at the center frequency when the horizontal width was 2.0 mm, as indicated by the solid red line, yielding excellent reflection coefficient characteristics. These results indicate that careful fabrication of grid cavities by accounting for their related factors is essential.
Fig. 7(c) compares the reflection coefficients attained for variations in the vertical width of the grid cavity. Iterative calculations were performed to determine the optimal vertical width, with the grid cavity height and horizontal width fixed at 5.0 mm and 2.0 mm, respectively. The vertical width was observed to influence the real part of the impedance, resulting in minimal frequency shifts and changes in the reflection coefficients. When the vertical width was 2.5 mm, as represented by the solid red line, excellent reflection coefficient characteristics below −40 dB were observed at the center frequency. Table 3 summarizes the optimal parameters for the simulated grid cavity.
Fig. 8 presents the structural diagram and calculated reflection coefficients for variations in the position of the grid cavity along the X, Y, and Z and the X-Y-Z axes. This evaluation assessed potential positional errors that may occur when fixing the grid cavity to the radiating antenna. The solid red line in Fig. 8(b) represents the case where X = Y = Z = 0.0 mm, indicating that the grid cavity had been correctly positioned without any error. Fig. 8(b) also confirms that positional errors at a fixed location in the grid cavity can shift the resonant frequency, which means that the overall impedance of the antenna is highly sensitive to positional inaccuracies in the grid cavity. Therefore, it is crucial to carefully fix the grid cavity when installing it on the radiating antenna since it has a drastic effect on the resonant frequency.
2. Analysis of Reflection Coefficients
Fig. 9 compares the reflection coefficients of the proposed antenna with and without the grid cavity. The blue dotted and solid red lines represent the simulation results for the antenna without and with the grid cavity, while the black dashed line indicates the measurement results of the reference antenna for comparison. Despite the addition of the grid cavity, the results indicate a good match between the resonant frequency point and the frequency bandwidth. This outcome can be attributed to the accurate design of the optimal grid cavity parameters and precise alignment of the combined position without any errors.
3. Analysis of Surface Current Distribution
Fig. 10 compares snapshots of the simulated surface current distributions of the antennas at the center frequency of 17.5 GHz. Fig. 10(a) and 10(b) show the surface current distributions of the antenna without and with the grid cavity, respectively, with the latter exhibiting a more concentrated surface current distribution than the former. This improvement is caused by the grid cavity minimizing impedance variations and suppressing the diffraction field near the edges of the radiating antenna. Therefore, the inclusion of the grid cavity is expected to result in a higher antenna gain.
4. Analysis of Radiation Patterns
Fig. 11(a) and 11(b) illustrate the simulated electric field radiation patterns Eθ and Eφ at 17.5 GHz, respectively. The blue dotted and solid red lines represent the simulation results for the antennas without and with the grid cavity, respectively, while the black dashed line denotes the measurement results of the reference antenna for comparison. For Eθ, the beam patterns of the antennas are similar, with the HPBW being approximately 10.45°. The first sidelobe of the antenna equipped with the grid cavity is suppressed by approximately 2 dB, suggesting a higher expected gain. Additionally, compared to the reference antenna, a significant overall improvement in sidelobe levels is observed. In the case of Eφ as well, the beam patterns are similar, with the HPBW being approximately 2.2°. Compared to the reference antenna, a significant overall improvement in sidelobe levels is observed. Additionally, the back lobe of the antenna with the grid cavity is suppressed by approximately 5 dB.
5. Analysis of Peak Gains
Fig. 12 compares the simulated peak gains of the proposed antenna. The blue dotted and solid red lines represent the simulation results for the antenna without and with the grid cavity, respectively, while the black dashed line denotes the measurement results of the reference antenna for comparison. At the center frequency, the antenna with the grid cavity exhibits a gain increase of approximately 1.62 dB and 2.93 dB over the antenna without the grid cavity and the reference antenna, as indicated by the blue dotted line and black dashed line, respectively. This improvement in gain can be attributed to minimal changes in impedance and radiation patterns despite the addition of the grid cavity, along with significant sidelobe suppression.
Based on these simulation results, a grid cavity and a 6 × 26 slotted waveguide array antenna were fabricated and combined.
IV. Fabrication and Measurement
Fig. 13(a), 13(b), and 13(c) display photographs of the fabricated radiating waveguide antenna, grid cavity, and the antenna combined with the grid cavity, respectively. Notably, the electrical dimensions of the fabricated antenna were 100 mm × 380 mm × 22 mm.
Fig. 14(a) depicts the interior of the anechoic chamber employed for antenna measurement, with the antenna under test (AUT) installed inside. Notably, this anechoic chamber supports a measurable frequency bandwidth ranging from 400 MHz to 24.5 GHz [28]. Fig. 14(b) and 14(c) illustrate the antenna setup for the AUT measurement. The yellow arrow indicates the rotation direction of the positioner for measuring the Eθ and Eφ electric field radiation patterns.
1. Measurement of Reflection Coefficients
Fig. 15 compares the reflection coefficients of the proposed 6 × 26 slotted waveguide array antenna featuring the grid cavity. The blue dotted line represents the simulation results of the proposed antenna, while the solid red line corresponds to the measurement results. The black dashed line denotes the measurement results of the reference antenna for comparison. It is observed that the simulation and measurement results are in good agreement. Given that the radar system required a frequency bandwidth of 500 MHz, or 2.8%, the measured frequency bandwidth of approximately 2.8% based on the −10 dB standard corresponded closely with the required bandwidth. This agreement can be attributed to precise adjustments of the height, width, and positional alignment of the grid cavity, as detailed in Section III.
2. Measurement of Radiation Patterns
Figs. 16 and 17 present a comparison of the electric field radiation patterns—Eθ and Eφ—at three frequency bands for the antenna combined with the grid cavity, as measured in the anechoic chamber. Notably, the measured radiation patterns were normalized.
Fig. 16 presents the radiation patterns measured for Eθ at each frequency. Owing to the characteristics of wave monitoring radar systems, which accurately measure the height of vertically moving nonlinear waves in real time, the elevation pattern (Eθ) of the antenna was designed to be significantly broad. The measured sidelobes and HPBW levels are summarized in Table 4.
Fig. 17 illustrates the measured radiation patterns of Eφ at each frequency. Based on the characteristics of a wave-monitoring radar system, the antenna was designed to have a narrower azimuth pattern (Eφ) compared to the elevation pattern (Eθ), since it had been designed to undergo full 360° mechanically rotates. Notably, a narrow azimuth pattern results in sharper directivity characteristics, thereby enabling a higher gain. The simulation and measurement results in Fig. 17 exhibit good agreement. Table 4 compares the sidelobe levels and HPBW results for the simulated and measured radiation patterns.
3. Measurement of Peak Gains
Fig. 18 compares the simulated and measured peak gains at 50 MHz intervals between 17.25 GHz and 17.75 GHz. The blue dotted and red solid lines represent the simulation and measurement results of the proposed antenna, respectively, while the black dashed line indicates the measurement results of the reference antenna for comparison. The simulated and measured gains of the proposed antenna show good agreement, with the measured peak gain at the center frequency of 17.5 GHz being approximately 30.5 dBi. This enhancement in antenna gain compared to the gain reported in [25] can be attributed to minimal changes in impedance and radiation patterns caused by the integration of the grid cavity. Furthermore, the improvement in the first sidelobe level can also be attributed to the incorporation of the grid cavity. It must be noted that a gain improvement of approximately 2 dB corresponds to a nearly 40% enhancement in power efficiency for the radar system, meaning that the results obtained in this study are highly significant.
V. Conclusion
In this paper, a Ku-band radar antenna for real-time signal processing of big data on waves is proposed so as to achieve a better performance than conventional X-band radars. A grid cavity was integrated into the radiating waveguide of a 6 × 26 slotted waveguide array antenna to achieve high gain, narrow beamwidth, and low sidelobes for the Ku-band radar antenna. A conventional grid cavity was redesigned and integrated with the radiating waveguide. Consequently, impedance variations occurred owing to design errors introduced during the assembly process after fabrication. As a result, some discrepancies between the measured and simulated reflection coefficients were identified. To address these shortcomings, the grid cavity and radiating element were both optimally designed, accounting for the errors that could occur during antenna fabrication. Subsequently, the mutual impedance effects were analyzed. The proposed approach not only improved the reflection coefficients, which were calculated after assembling the grid cavity and the antenna during fabrication, but also enhanced the agreement between the simulation and measurement results.
Additionally, an antenna equipped with the proposed grid cavity was designed and fabricated, and its performance was analyzed. The proposed antenna exhibited a bandwidth of approximately 2.8% at the −10 dB level, thus fulfilling the frequency bandwidth requirements for wave-monitoring radar systems. The measured radiation patterns closely matched the simulation results. Furthermore, the measured HPBW of Eθ at the center frequency was 10.05°, indicating a broad elevation angle, while the sidelobe level of Eθ improved by approximately 2 dB upon adding the grid cavity. The measured HPBW of Eφ at the center frequency was 1.98°, reflecting a narrow azimuth angle. The measured peak gain of the antenna was closely aligned with the simulated peak gain results, exceeding 30.5 dBi at the center frequency. Notably, compared to the antenna without the grid cavity [25], the gain increased by approximately 2 dB. Overall, these results imply that the proposed antenna can contribute to the advancement and development of wave-monitoring radar systems for coastal erosion prevention.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
Print
