A Compact All-Band Spacecraft Antenna with Stable Gain for Multi-Band GNSS Applications

Abstract

This study presents a compact and stable gain spacecraft antenna that operates in all Global Navigation Satellite System (GNSS) bands from 1.164 GHz to 1.610 GHz. The proposed antenna structure based on the single-feed crossed bowtie antenna concept consists of four triangular patches excited with a 90° phase difference in between to generate right-hand circular polarization (RHCP), without needing complex feed networks. The radiator part of the antenna is covered by a radome and is also supported by a cylindrical dielectric cavity frame (DCF) to weaken the diffracted waves propagating along the ground plane while increasing vibration resistance. The fabricated antenna provides a return loss better than 10 dB with lower than 3 dB axial ratio and a stable gain around 7.2 ± 0.3 dBic over the entire GNSS bands, as well as a more compact and lightweight structural performance. It is also verified that the structural integrity and functional performance of the fabricated antenna remain consistent despite exposure to an equivalent vibration level in the launch process. The presented all-band spacecraft GNSS antenna is an innovative implementation with space industry insight for multi-band space applications that have application-specific limitations and provides consistent performance, as well as operational safety with the antenna design simplicity.

1. Introduction

Global Navigation Satellite Systems (GNSSs) are described as satellite constellations strategically placed to provide positioning, navigation, and timing (PNT) data. Specifically, the systems called Galileo, Compass, GLONASS, and GPS are the four most well-known PNT providers, and each of them has provider-specific operational frequency bands from 1.164 to 1.610 GHz [1]. This diversity in GNSSs increases the demand for advanced GNSS receiver systems using different frequency bands of various GNSSs. The multi-band receivers improve the positioning accuracy and reliability while reducing jamming vulnerability of the system [2]. The all-band antennas are the indispensable parts of multi-band GNSS receivers to maximize the operational capability by receiving signals from all operating bands provided by each GNSS using just a single antenna. Thus, the number of antennas, interference problems between antennas, and the antenna placement difficulties can be easily reduced. This is especially advantageous for space platforms that are subject to severe size and weight limitations. Moreover, if the all-band antenna has a stable gain performance over the entire frequency band, it provides another significant advantage by reducing the design complexity of the amplifier network used in multi-mode GNSS receivers.

GNSSs operate in circular polarization to mitigate ionospheric losses, and receiving GNSS antennas are typically designed to provide right-hand circularly polarized (RHCP) radiation characteristics to decrease the polarization mismatch to as low a level as possible in order to increase the efficiency of data transmission as high as possible [3]. Particularly for space applications where weight and size are strictly limited, a good candidate for an all-band GNSS antenna must have high-compactness while ensuring its performance remains constant under all operational frequency bands. However, theoretically the fundamental limitation of antenna bandwidth is directly related with its electrical size [4]. These two conflicting requirements are the major challenge of all-band GNSS antenna designs for applications where size and weight are limited including space platforms.

A wide variety of antennas are used in GNSS receiver studies, but most of these types are not suitable for direct utilization in practical space implementations. One of the candidates is quadrifilar helix antennas (QHA) because of their satisfactory axial ratio performance in circular polarization, whereas these antennas operate in resonant modes and have inherently narrow bandwidth [5]. Recently, some new versions of QHA have been presented that are aimed at obtaining wider bandwidth or multi-band behavior [6,7,8]. However, they require an additional broadband four-way hybrid feed to cover all GNSS bands, which complicates their feeding networks. Additionally, QHA needs a relatively large ground plane to provide the expected high gain performance. Patch antennas are another popular type of antenna that are easy to manufacture and offer a low-volume (planar) option for some GNSS applications [9]. Nevertheless, they are also resonant structures with limited bandwidth, and various methods have been presented to overcome their narrow-band characteristics, such as stacked structure [10], U-shaped slot [11], and L-probe [12]. The designed structures using these bandwidth enhancement methods require additional complex feed networks and are quite poor in resistance to vibration, which is vital in space applications. Although the operational bandwidth is not wide enough to operate in all GNSS bands, another study that is suitable for utilizing in space applications is on the dielectric resonator antenna (DRA) type [13]. Thanks to its monolithic structure, vibration resistance is sufficient, but it only has satisfactory performance in the lower and upper GNSS bands. Another space application-compatible antenna study presents dual/triple band designs and an antenna array for GPS [14]. Although the antennas presented in the study are not able to cover all GNSS bands, their designs are suitable for the thermal and vibrational constraints of space applications. One particular well-known solution that provides a full GNSS band coverage performance is the four-element ring array with different element types, such as Vivaldi, F/inverted-F, or triangle patch [15,16]. Also, the performance of such antennas can be increased by implementing a metal cavity. However, metal cavities dramatically increase the overall weight of the antenna, which is undesirable for space applications [17].

The antenna type utilized in the design of this study is principally based on orthogonally placed symmetric crossed bowtie dipole antennas [15,16,17,18,19]. Although more complex antenna designs with asymmetric dipole configurations or with an additional metal cavity structure are presented in the literature utilizing this basic structure, the simplest form of the antenna type is preferred in this study due to the need for high simplicity and reliability in space applications [18,19]. On the other hand, the additional requirements of a spacecraft antenna due to the application environment and operating conditions such as high robustness, radome protection, compactness, lightweight, thermal resistance, and low outgassing material usage are not considered in these studies in the literature, and the presented designs are structurally vulnerable. A more comprehensive spacecraft antenna is proposed in this study, considering the application-specific requirements that dramatically affect operational performance of the antenna design.

This study presents a single-feed, stable gain, RHCP-crossed bowtie antenna covering all GNSS bands, with a compact size and lightweight structure. The proposed antenna structure is reinforced with a dielectric cavity frame (DCF) to mitigate diffracted waves while enhancing its robustness. While the preliminary prototypes previously published in [20,21,22] also pursued to meet the same all-band GNSS antenna goal needed in multi-band operations of space applications, the proposed final design in this study provides an improved and more stable operational performance in terms of gain, axial ratio, and cross-polarization, as well as a more compact and lightweight structural performance. Moreover, by evolving a well-known simple and delicate antenna type with space industry insight, it paves the way for its utilization in space applications and offers a strong GNSS antenna candidate for space vehicles.

2. Antenna Design

The proposed antenna design is composed of four main parts as radiator, conducting ground plane, DCF, and radome. In order to utilize the design efficiently in space applications, the radiator part should be protected by a low-loss radome that is also resistant to temperature changes. Additionally, the radiator needs to be supported by a lightweight structure around the edge so as not to reduce the ground plane effect and to withstand the vibration caused by the launch process. Accordingly, a dielectric cavity support structure and radome are designed by using Ketron 1000 PEEK (ε_r = 3.2, tanδ = 0.002) material considering its distinguishing features for space applications such as high robustness, lightweight, low dielectric loss, and low outgassing. In addition, the thermal resistance of selected material is quite satisfactory enough to meet the need for the structural integrity of the antenna to remain stable during highly variable thermal conditions in the space environment. Moreover, the DCF and radome are also mounted to each other with screws specially fabricated from PEEK in order to increase structural strength and prevent possible structural integrity deterioration that may arise from the difference in thermal expansion of the used materials.

The radiator part is built on a RO4003C (ε_r = 3.38, tanδ = 0.0021) substrate with 0.813 mm thickness. Its design consists of two bow-tie dipole antennas (in other words, four triangular patch elements). Two of the patch elements placed as perpendicular to each other on the top surface of the substrate are shown in Figure 1a with design parameters. The other two elements of the radiator part are placed on the bottom surface as a mirror image symmetric of their counterparts on the top surface, as given in Figure 1b.

The planar base surface of the antenna structure is a conductor that acts as a ground plane. A metallic hollow mast is physically used to stabilize the clearance between the radiator part and the ground surface in a vibration-proof manner, as shown in Figure 2. In addition, a 50 Ω coaxial cable, which is utilized to feed the triangular patch elements on the radiator part, also passes through the metallic mast. In addition, the outer conductor of the coaxial cable is connected to both this metallic mast and the patch elements on the bottom surface, while the inner conductor is connected to the central feed point of the patch elements on the top surface.

Another important issue to be examined in antenna designs fed by a coaxial cable is the feeding efficiency losses that may arise from the unbalanced phase nature of the coaxial cable and the possible impedance mismatch between the radiator and the coaxial cable. In the antenna design in this study, the unbalance and mismatch effects of the 50 Ω coaxial feed are analyzed with an ideal delta-gap feeding (Lumped Port in HFSS or Discrete Port in CST Studio) assigned between the central feed points of the patch elements on the top and bottom surface of the substrate. Thus, the radiator part of the antenna is fed by an ideal feed by eliminating the effect of the coaxial cable, and the reflection coefficient and axial ratio results are compared with the practically applied coaxial feed case, as given in Figure 3. The comparative results show that the proposed antenna design is quite suitable for coaxial feeding.

In order to generate an RHCP, the vertically placed patch elements on the top and bottom surface are respectively fed by the inner and outer conductor of the coaxial cable, which inherently have a 180° phase difference between of them. The horizontally placed patch elements on each surface are connected to vertically placed counterparts by ring-shaped delay lines that provide two additional 90° phase delays. From another perspective, the proposed radiator can be considered integrated, as two identical bow-tie dipole antennas are placed on the x and y axes and with a 90° phase difference between the feedings of them. Therefore, a circularly polarized wave is produced that has an electric field vector that revolves on the x-y plane. Since the triangular patch elements are identical,

$$\left|{\stackrel{⃑}{E}}_x\right|=\left|{\stackrel{⃑}{E}}_y\right|,$$

(1)

and the obtained RHCP wave can be expressed as

$$\stackrel{⃑}{E}\left(z,t\right)={\stackrel{⃑}{E}}_x\cos \left(wt-kz\right)+{\stackrel{⃑}{E}}_y\cos \left(wt-kz+{\displaystyle \frac{\pi }{2}}\right).$$

(2)

Thus, the RHCP wave is achieved with an onboard solution using just a single-feed. Therefore, the proposed antenna design does not need any extra hybrid couplers or any complex feed network, which adversely affects the reliability of the antenna performance in the space environment. On the other hand, even if the greater part of the RHCP wave is reflected by the ground plane, diffracted waves can still propagate across the plane and increase the back lobe of the radiation pattern [17]. Therefore, it is aimed to weaken the diffracted waves by changing the boundary conditions with a relatively thick dielectric cavity frame placed perpendicular to the reflective ground plane. In order to observe this effect of DCF on back lobe radiation, the proposed antenna structure with DCF is simulated without DCF for the starting and ending frequencies of the operational frequency band, and the radiation pattern results are compared, as shown in Figure 4. It is observed that the use of DCF in the proposed antenna design prevents the constructive effect of the diffracted waves on the back lobe by scattering and helps to weaken them, while increasing the vibration and thermal resistance.

The antenna structure is designed and optimized by utilizing ANSYS HFSS. The designed antenna is presented in Figure 5, including details such as mounting apparatus, DCF, radiator, and radome to make the simulation conditions as close as possible to the measurement conditions. Thus, all possible factors that are presumed to affect antenna performance are fully simulated and optimized before the fabrication and measurement stage.

The initial values for the optimization of the designed antenna are determined by experience gained from the simulation and measurement results of the preliminary prototypes of this study, previously presented in [20,21,22]. The ultimate values of the optimized GNSS antenna design parameters are determined by taking the fabrication precision and capabilities into account. The manufacturable ones are evaluated as the final design parameter values of the proposed antenna structure. The final production parameters determined for the presented space vehicle compatible antenna structure are given in

Table 1. Antenna design parameters.

Parameter	Value	Description
Rsub	64.6 mm	substrate radius
L	42.5 mm	length of triangular patch
α	80°	triangular patch angle
Ri	7.4 mm	inner radius of phase delay line
Wi	1.6 mm	phase delay line width
H	43 mm	radiator height from the ground
Rmast	4.25 mm	cylindrical mast outer radius
Rcoax_diel	1.54 mm	outer radius of coaxial cable dielectric
Rcoax_in	0.46 mm	radius of coaxial cable inner conductor
TDCF	8 mm	wall thickness of DCF
Hradome	6.15 mm	air gap from radiator to radome
Tradome	1.1 mm	radome thickness

.

The radius of the ground plane is limited by the radius of the radiator substrate (R_sub) and, hence, the length of triangular patches (L). In order to achieve a compact antenna design, the radius of the antenna structure should be as close as possible to the radiator width, dominantly determined by the L value, which is directly dependent on the operating frequency.

3. Antenna Fabrication and Test

The designed antenna is implemented on a metallic reflective ground plane fabricated with a monolithically hollow cylindrical mast to increase the vibration resistance of the structure. The radiator part is fixed to the DCF with PEEK screws, and the inner conductor of coaxial feed is connected to the center of the radiator, as shown in Figure 6a. The monolithic nature of the ground plane with the mast is critical in terms of structural deterioration problems that may occur due to the potential for sudden flexing between the side walls where the radiator substrate is mounted and the central feeding connection point, which is the weakest point of the structure against vibration. On the other hand, the monolithic cylindrical mast provides circumferential support by soldering its outer perimeter to the bottom surface of the radiator and increases the vibration strength of the central feeding connection in its perforated interior, as shown in Figure 6b. In addition, choosing the substrate that is compatible with deformation-free flexing as much as possible reduces the transmission of vibration within the structure and reduces the possibility of structural fracture. The assembly process is completed by a radome made of the same PEEK material with a thickness of T_radome. The total weight of the built-to-launch structure is 493 gr, the diameter is 138.8 mm, and the height is 60.8 mm. The achieved overall volume reduction is 35%, and the weight reduction is 30% compared to the second prototype presented in [21,22], and the achieved structural reductions increase when compared to the first prototype presented in [20], as 57% and 48%, respectively.

The reflection coefficients of the fabricated antenna versus the frequency are measured and compared with the simulation results, as shown in Figure 7. The fabricated antenna provides a return loss of better than 10 dB in all GNSS bands and less than 15 dB in most of these bands. The results demonstrate that the antenna presents a very satisfactory return loss performance from 1164 MHz to 1610 MHz and provides a very efficient operating performance covering all GNSS bands. This means that in all bands, at least 90% of the antenna’s input power, which is very limited in space applications since the power is obtained through solar panels and batteries, is transmitted as radiation despite all practical material losses.

SATIMO StarLab is used in the antenna radiation performance, including axial ratio and co- and cross-polarization radiation pattern measurements. The utilized measurement system for the radiation pattern performance is shown in Figure 8.

The proposed antenna is exposed to variable vibration levels standardized by the European Cooperation for Space Standardization (ECSS), and its operational performance is examined in terms of vibration resistance [23]. The utilized measurement system for vibration (sine, random, and shock) tests is given in Figure 9.

The radiation performance measurements are repeated after the vibration test (AV) to verify that the manufactured antenna maintains its structural integrity when exposed to vibration and to observe whether vibration affects the radiation performance of the antenna. Both measurement results that are obtained before and after the vibration tests for the gain, cross-polarization, and axial-ratio are compared with the simulation results, as illustrated in Figure 10, Figure 11, and Figure 12, respectively.

Considering the operating conditions of the proposed GNSS antenna in the space environment, the antenna structure needs to be resistant to vibrations caused by launch vehicles and temperature changes due to orbital motion. The produced antenna structure successfully passed visual and X-ray inspections to observe possible structural deformations under high-intensity vibration conditions. On the other hand, the proposed DCF structure for spacecraft antennas was also implemented on IMECE, the first earth observation satellite of Turkiye with sub-meter resolution, and its thermal resistance is also verified operationally [24].

In spite of all the challenging test conditions, the proposed GNSS antenna provides approximately 7.2 dBic gain with a maximum ±0.3 dB fluctuation observed in all GNSS frequency bands with an axial ratio better than 3 dB. Also, the cross-polarization level, i.e., relative magnitude level of the left-hand circular polarization (LHCP), is less than −13 dBic, even in the worst case, for the whole GNSS bands from 1.164 to 1.610 GHz. In addition, the center frequencies of E1 and E5a bands of Galileo, L1, L2 and L5a bands of GPS, and G1 band of GLONASS are selected to observe the co- and cross-polarization radiation patterns of the fabricated antenna. The measurement results for two principal cut-planes are also shown in Figure 13 and Figure 14. Evaluation of the pattern measurement results indicates that the antenna has a very satisfactory radiation pattern with a wide angular range. As a result of this feature, the proposed antenna is not only able to receive signals from a larger number of GNSS providers but also significantly increases the signal reception time window with each of them. Thus, the positioning accuracy and reliability of the spacecraft are significantly increased by using only a single compact and lightweight antenna.

The proposed antenna is also a strong candidate when compared with existing GNSS antennas for space launch vehicles. A comparable spacecraft compatible GNSS antenna study is the patch excited cup (PEC) antenna for precise orbit determination of low Earth orbit (LEO) satellites [25,26]. This PEC antenna also has very similar dimensions with the antenna proposed in this study. However, it does not support all-band GNSS operations and only has two operational frequency bands, L1 at 1.575 GHz and the combined L2 plus L5/E5a at 1.227 and 1.176 GHz. Although these operating frequency bands cover the entire GPS bands and allow use in multi-frequency applications, they are not suitable for applications with multiple different PNT providers. On the other hand, the results of the antenna proposed in this study show that it is not only compatible with multi-frequency applications but also is suitable to operate this application with different GNSS operators, such as Galileo, Compass, GLONASS, and GPS.

4. Conclusions

Antennas providing stable gain performance in all GNSS operational frequency bands are highly critical elements for practical space applications that require precise location determination. The capability to receive data from multiple GNSS providers with only a compact wideband antenna, compared to using multiple antennas for different GNSS bands, provides advantages such as elimination of the interaction problems between antennas, minimizing the volume occupied by the antennas on the satellite, and reducing the total platform weight, which are critical for space applications. In this study, a space vehicle compatible stable gain GNSS antenna operating in all GNSS frequency bands is presented. The proposed antenna provides a return loss better than 10 dB, an axial ratio lower than 3 dB, a stable RHCP gain higher than 7.2 ± 0.3 dBic, and a low cross-polarization level over the entire frequency bands of GNSSs. The achieved overall volume reduction is 35%, and the weight reduction is 30%, compared to the second and third prototypes presented in [19,20] and compared to the first prototype presented in [18], as 57% and 48%, respectively. The vibration resistance of the fabricated antenna is verified to qualification-level vibration testing, and the structural integrity and high functional performance of the antenna remain consistent. According to results, the fabricated and measured antenna is a very strong candidate for use in multi-band GNSS implementations for satellite applications, thanks to its simple design and satisfactory wideband performance. Moreover, the functional capability of the proposed new antenna structure and its structural resistance to environmental conditions of space applications have been operationally verified with a real spacecraft currently performing its mission.