WO2021034270A1

WO2021034270A1 - A linear-to-circular polarizer, feeding network, antenna and antenna assembly

Info

Publication number: WO2021034270A1
Application number: PCT/SG2020/050473
Authority: WO
Inventors: Yongxin Guo; Jie Wu; Cong Wang
Original assignee: National University Of Singapore
Priority date: 2019-08-16
Filing date: 2020-08-14
Publication date: 2021-02-25

Abstract

Disclosed is a linear-to-circular polarizer comprising a cavity and a pair of stepped ridges. The stepped ridges are disposed in the cavity and define a gap therebetween. Also disclosed is a feeding network for combining four CP elements into a 2 by 2 array antenna, comprising a first stage power divider and a second stage power divider.

Description

A LINEAR-TO-CIRCULAR POLARIZER, FEEDING NETWORK, ANTENNA

AND ANTENNA ASSEMBLY

Technical Field

The present invention relates, in general terms, to a linear-to-circular polarizer and a waveguide circularly polarized element antenna including such a polarizer. The present invention also relates to a feeding network and an array antenna assembly including such a network and polarizer.

Background

Circularly polarized (CP) antennas, have been widely used in many wireless applications such as satellite communications and navigation due to their superiority in the polarization matching. In the over-the-horizon communications and navigation systems, CP array antennas are good candidates to provide high gain and good resolution.

In recent years, work has focused on improving the working bandwidth and efficiency of CP array antennas, being properties that are essential for realizing frequency diversity or reduced energy consumption of communications and navigation systems. The working bandwidth of a CP array antenna is mainly determined by the overlapping or overlapped bandwidth of impedance matching and qualified axial ratio (AR). For specific applications, there may also be limitations relating to gain fluctuation and pattern deformation.

Generally speaking, there are mainly three categories of methods to implement CP arrays. CP beams can be achieved at the CP elements, by sequentially fed networks or by using additional polarizers. There are some other solutions, such as the radial line slot antenna which is efficient except that the bandwidth is too narrow for most applications. Among the mainstream approaches, a CP array with CP elements assembled by a full corporate feeding network is more compact and efficient when compared with other solutions, but the bandwidth of such a topology needs to be improved. Some innovations have been explored to expand the working bandwidth of the CP element for some CP arrays. For example, aperture-coupled magneto-electric dipoles fed by a substrate integrated waveguide (SIW) feeding network have been proposed - in some instances, these arrangements involve stacked curl elements utilized as the CP radiating elements in a CP array fed with a corporate SIW feeding network. Such arrangements can result in impressive bandwidth improvements. However, the feeding network formed by SIW can lead to a drastic degeneration of antenna efficiency when the array scale goes beyond 16 x 16.

The total efficiency of a CP array is mainly determined by the insertion loss of its feeding network, which is primarily affected by the array scale and the transmission lines used. For a small array, the efficiency is not so sensitive to the type of the transmission lines. But things are quite different for large arrays. Transmission lines are often a primary consideration before designing a high- gain and high-efficiency array antenna.

Waveguides have smaller insertion loss than the other competitors in microwave and millimetre-wave applications, and have been widely used as the feed line in many large-scale arrays to improve the antenna efficiency. However, the dimensions of the waveguide apertures are usually larger than most planar radiating elements such as micro strip patches. Thus, it is difficult to directly combine all the waveguide elements with a simple parallel feeding network due to the very limited spacing between adjacent radiating elements. To solve the layout problem, higher order mode cavities are commonly utilized to feed a 2 x 2 subarray. When the feeding problem of the 2 x 2 subarray has been addressed, there will be sufficient spacing to implement an expanded feeding network for larger-scale arrays. This solution works well except that the working bandwidth of the array is inevitably limited by the relatively narrow bandwidth of the high-order mode cavity which is used as the feeding structure of the 2 x 2 subarray. Another method of feeding a high-efficiency array antenna has been proposed, in which a corporate gap waveguide feeding network is used to feed chamfered cylindrical apertures, but the working bandwidth is limited by the radiating elements. It would be desirable to overcome or ameliorate at least one of the above- described problems with existing CP antenna assemblies, or at least to provide a useful alternative.

Summary

Proposed herein are novel wideband high-efficiency circularly polarized (CP) waveguide array antenna arrangements for wireless communications. In many embodiments, two key considerations were taken into account when determining how to construct a large-scale waveguide CP array antenna with wide bandwidth and high efficiency. These two considerations are: how to design wideband CP elements; and, how to design a wideband and compact feeding network for the basic subarray.

Embodiments of the present invention can be fabricated using the 3D printing technology, and can be further expanded to larger-scale arrays while keeping very high efficiency and wide bandwidth

Disclosed herein is a linear-to-circular polarizer comprising: a cavity; and a pair of stepped ridges disposed in the cavity, the stepped ridges defining a gap therebetween.

The stepped ridges may be disposed on opposite sides of the cavity. The stepped ridges may, for example, be antipodally disposed.

The polarizer may further comprise a feeding port and a radiating port, the polarizer polarizing a linearly polarized wave, received from the feeding port, to a circularly polarized wave at the radiating port after propagating through the polarizer. The radiating port may be a square aperture.

The stepped ridges may taper towards the radiating port. The gap may increase towards the radiating port.

The polarizer may further comprise a pair of stepped matching blocks disposed in the cavity, the stepped matching blocks being spaced from the stepped ridges and defining a gap therebetween. The stepped matching blocks may be disposed on opposite sides of the cavity.

The stepped matching blocks may taper towards the radiating port.

The stepped matching blocks may be laterally offset and each matching block extends into the cavity, starting from the feeding port.

Also disclosed is a waveguide circularly polarized (CP) element antenna comprising: a) a feeding waveport; b) a linear-to-circular polarizer as described above; and c) a radiator for radiating the circularly polarized electromagnetic wave.

Also disclosed herein is a feeding network for combining four CP elements into a square array antenna, comprising a first stage power divider and a second stage power divider. The square array antenna may be a 2 x 2 antenna.

The first stage power divider may be a l-to-2 power divider, and the second stage power divider is a 2-to-4 power divider. The l-to-2 power divider may be an E-plane power divider and the 2-to-4 power divider may be composed of two H-plane power dividers. The feeding network may further comprise a first septum and a second septum inserted into a multiple stage cavity to form the first stage power divider and the second stage power divider. The multiple stage cavity may have a smaller open end and a larger open end, the larger open end being larger than the small open end.

The first septum may be orthogonally crossed over the second septum, extending from the larger open end toward the smaller open end to form one input port at the smaller open end and four output ports at the larger open end, both septa with different dimensions touching lateral walls of the multiple stage cavity.

The first septum may have a consistent thickness and the second septum may have a first thickness and a region of a second thickness, thicker than the first thickness.

Also disclosed herein is an array antenna assembly comprising: four waveguide CP element antennas, each waveguide CP element antenna being a waveguide antenna as described above; and a feeding network as described above, the output ports being connected to the feeding ports of all four waveguide CP element antennas by a waveguide transition.

Advantageously, embodiments of antennas proposed herein provide wide bandwidth and high efficiency. Thus, such embodiments can be used in many wireless communication systems that rely on frequency diversity or low/reduced energy consumption.

Advantageously, some embodiments employ a fully metallic structure. The fully metallic structure is an excellent candidate for space applications such as satellite communications. Waveguide CP array antennas have many applications due to their superiority in the polarization matching and efficiency. But there are some layout problems for the basic subarray. High order mode cavities are used to solve such problems in many previous works, but the antenna bandwidth is inevitably limited by the high order modes used. Advantageously, an approach described herein, for designing a compact 2x2 waveguide CP array, results in an array that demonstrates significantly improved bandwidth (around 40.3% improvement) and a total efficiency higher than 80%. Advantageously, designs formed in accordance with such an approach can be sufficiently compact to hide the feeding network underneath the radiating apertures, and can be readily expanded to larger-scale arrays while maintaining very high efficiency and almost equivalent bandwidth.

Brief description of the drawings

Embodiments of the present invention will now be described, by way of non limiting example, with reference to the drawings in which:

Figure 1 illustrates an embodiment of circularly polarised (CP) array antenna in accordance with present teachings;

Figure 2 illustrates geometry of an embodiment of an antipodally ridged circularly polarised element in accordance with present teachings, in which Figure 2(a) is a cross-section through the yz plane, Figure 2(b) is a cross-section through the xz plane, and Figure 2(c) illustrates a stepped ridge;

Figure 3 illustrates the operational principle of the invention analysed using a transmission line analysis method, in which Figure 3(a) shows excited modes: Port 1 with TEio mode; Port 2 with TEio mode; and Port 3 with TEoi mode, Figure 3(b) shows the amplitude and phase imbalances of the energy E shields at Port

2 and Port 3, and Figure 3(c) graphically illustrates the calculated axial ratios over frequencies using Equation (1). Figure 4 shows the simulated performance of the antipodal ridged CP element, in which Figure 4(a) shows the simulated reflection coefficient over a gigahertz (GHz) frequency spectrum, and Figure 4(b) shows the simulated realised gain over the same spectrum;

Figure 5 illustrates the parametric analysis of amplitude imbalances and phase differences between Port 2 and Port 3 affected by hi using the transmission line model illustrated in Figure 3;

Figure 6 shows a comparison of axial ratios between CP antennas with linear ridges and stepped ridges, in both cases the designs comprising dual ridges antipodally disposed in a multistage cavity; Figure 7 illustrates a consideration of the compact feeding network in accordance with present teachings, in which Figure 7(a) shows a cross-section in the yz plane, and Figure 7(b) shows a cross-section in the xz plane;

Figure 8 shows the simulated reflection coefficients of the input port (Port 1) and power amplitudes of the output ports (Port 2, 3, 4 and 5) of the proposed feeding network;

Figure 9 is a photograph of the fabricated (2 x 2) CP array antenna; Figure 10 is the simulated and measured reflection coefficients of the proposed 2x2 CP array antenna;

Figure 11 illustrates the simulated and measured gains and axial ratios of the fabricated CP array antenna, the illustrated gains being the realised right-hand circularly polarised (RHCP) gains; Figure 12 shows simulated and measured normalized RHCP and left-hand circularly polarised (LHCP) radiation patterns of a CP array antenna in accordance with present teachings, in which the measurements are shown for Figure 12(a) 10 GHz at the f = 0° plane, Figure 12(b) 10 GHz at the f = 90° plane, Figure 12(c) 12.3 GHz at the f = 0° plane, Figure 12(d) 12.3 GHz at the f = 90° plane, Figure 12(e) 14.6 GHz at the f = 0° plane, and Figure 12(f) 14.6 GHz at the f = 90° plane; and

Figure 13 illustrates an expanded 8x8 CP array antenna based on the proposed subarray, in which, Figure 13(a) shows the physical consideration of the CP array antenna, and Figure 13(b) provides the simulated performance characteristics.

Detailed description

Described with reference to the drawings is a wideband circularly polarized (CP) waveguide array antenna. Embodiments of the proposed array antenna consists of four antipodally ridged elements and a compact feeding network with two orthogonal septa inserted to a stepped cavity. Both components are wideband and they can work independently or together. The proposed topology gives a good solution to feed a 2x2 waveguide subarray and can form the basis of large- scale arrays. When compared with array antennas fed by high order mode cavities, the present design significantly improves performance and shows good potential for larger-scale arrays - e.g. square or rectangular arrays. Moreover, embodiments utilising a fully metallic structure are excellent candidates for satellite communication applications.

In some CP antennas, a pair of antipodal notches is introduced on the open end of the SIW to form a wideband CP antenna. Such a structure could sustain two orthogonal electric fields simultaneously and the relative amplitude and phase difference of the orthogonal electric fields can be flexibly controlled by adjusting the dimensions of the antipodal notches. Developing further on this concept, some embodiments herein provide a wideband CP element antenna with dual stepped ridges antipodally inserted in a square waveguide.

It is possible to achieve dual CP polarizations simultaneously using a ridged waveguide polarizer with a single ridge, but the bandwidth of the port-to-port isolation is difficult to improve using such a topology. Deteriorations in isolation adversely influence the AR performance at both channels. Proposed herein are antipodally placed ridge pairs for replacing the single ridge to improve the working bandwidth, but typically only one polarization is kept. As discussed below, the antipodal layout plays a vital role in sustaining components of the electromagnetic field with particular directional polarisation, and the stepped ridge in some cases is an optimized design derived from a linear shape.

Embodiments of the proposed antipodal ridged waveguides disclosed herein can be used as a compact and wideband CP radiator, and can also be used as a wideband waveguide polarizer with small insertion losses. As a consequence, a compact l-to-4 full corporate feeding network is also proposed herein, which can similarly be upscaled for larger arrays. Such a feeding network provides wider bandwidth than feeding networks formed using higher order mode cavities. In addition, the compact l-to-4 feeding network is used to assemble four antipodally ridged CP elements together to form a 2 x 2 CP array antenna. The feeding network and test port of some embodiments are completely hidden under the radiating apertures of the proposed CP array, as shown in Figure 1. Measured results demonstrate a working bandwidth of 40.3% from 9.9 GHz to 14.9 GHz.

Lastly, an extended 8 x 8 array antenna is also designed based on the previous subarray and exhibits a simulated total efficiency over 78% throughout the frequency band of 10-14.8 GHz. Figure 1 shows an array antenna in accordance with present teachings. The antenna 100 is a 2 x 2 CP array antenna comprising a 2 x 2 (1.64L0) array 102, dual ridged circularly polarising element 104, transition 106 and a compact 1- to-4 feeding network 108. Flanged connection 110 enables the antenna 100 to be connected to an external structure. At the base of the antenna 100 is a waveguide 112 (e.g. a WR75 waveguide or any other waveguide appropriately dimensioned for the required application and antenna). The dual ridged circularly polarising element 104 can be supplied separately, in some cases, though is presently described in the context of antenna 100 for illustration purposes.

The antenna 100 can be formed by using a transition to connect four antipodally ridged CP elements with the compact feeding network together, as shown in Figure 1. A validation prototype shown in Figure 9 is fabricated using 3D printing and milling technologies and will be discussed later. To improve ease of fabrication, the 2x2 CP array antenna 100, and other antennas formed in accordance with present teachings, can be separated into two or more parts. In the example shown in Figure 1, the antenna 100 may be formed as a single, unitary part or as two parts - e.g. Part I and Part II, that are processed using three-dimensional (3D) printing and milling technologies respectively. Notably, in the embodiment shown in Figure 1, Part I must be 3D printed, whereas 3D printing or milling may be used to produce Part II. Both parts are fabricated independently and then connected together by screws, welding, adhesion or any other appropriate method.

Linear-to-circular polarisation is achieved at polarizer 104. The polariser 104 includes a cavity 114 and a pair stepped ridges 116 disposed in the cavity. The stepped ridges define 116 a gap 118 therebetween. The dual stepped ridges 116 are disposed on opposite sides of the cavity 114 - i.e. ridges 116 are antipodally inserted in a square waveguide 120. The resulting device 100 can be used as a compact and wideband CP radiator (single-fed radiator), and may also be used as a wideband waveguide polarizer with small insertion losses.

With regard to the design of the CP element, in the CP element 200 of Figure 2, the cavity 204 is a multistage cavity. The two stepped ridges 202 are identical and are presently placed antipodally in the centre of the multistage cavity 204, like a tape slot antenna inserted into a square waveguide with two stepped impedance matching blocks - in fact, the present polariser 200 may include, in addition to the ridges 202, two stepped matching blocks 206 disposed in the cavity 204 and spaced from the stepped ridges 202 and defining a gap therebetween. The stepped matching blocks 206 are disposed on opposite sides of the cavity 202, and also taper (i.e. narrow) towards the radiating port (Part III). The stepped matching blocks 206 are laterally offset (i.e. to the sides in the view shown in Figure 2(a)) from each other and extend into the cavity starting from the feeding port (Part I) at which they are widest.

The CP element 200 can be decomposed into three modules as shown in Figure 2. Part I (see Figure 2(a)) functions as a feeding port or feeding waveport. Part II forms the polarizer transforming a linearly polarized (LP) electromagnetic field to a CP one. Part III is a radiating port or square radiator having square aperture 210. The stepped ridges 202 taper towards the radiating port (Part III), the gap 204 consequently increasing towards the radiating port (Part III). The polariser 104 polarises a linearly polarised wave, received from the feeding port (Part I), to a circularly polarised wave at the radiating port (Part III) after propagating through the polariser 104.

Thus, broadly proposed with reference to Figure 2, is a waveguide CP element antenna comprising a feeding wave port (Part I), a linear-to-circular polariser (Part II) and a radiator (Part III) for radiating a circularly polarised electromagnetic wave.

To illustrate where the design values of the proposed embodiment are located, cross-section view of one antipodally ridged CP element 200 is shown in Figure 2(c) with the design values listed thereon. The design values are also listed in Table 1. Only three key parameters are determined with reference to CP element 200 and thus a good result can be obtained using iterative parametric refinement alone. The linear shaped ridge can be further discretized into several parts to finally result in the stepped shape. While two or more steps may be used, the present stepped ridges comprise six steps, with the increasing number of steps enabling progressively more degrees of freedom in optimization, particularly when compared with a linear shaped ridge. Better performance of the present stepped ridges is therefore expected.

Table 1: proposed values for the dimensions shown in Figure 2 To illustrate the operational principle of CP elements designed in accordance with present teachings, a transmission line analysis method is adopted for analysing the CP element 300 shown in Figure 3(a). Port 1 has dimension of ao x hi (i.e. the width of the cavity or of the square radiating aperture x the width of the gap between stepped ridges/antipodal ridges). Only the transverse- electricio (TEio) mode is excited on Port 1 at the centre frequency. To monitor the mode conversions, another two ports i.e. Port 2 and Port 3, are added on the output of the CP element 300.

TEoi mode can be produced by introducing the antipodal ridge pair. This is because partial electromagnetic (E) fields with the /-polarization will be twisted into the x-polarization to satisfy the boundary conditions. By appropriate selection of the parametric values, most energy from the input port will be divided into two orthogonal modes, i.e. TEio mode and TE_0i mode at ports 2 and 3 respectively.

By analysing the transmission coefficients of S-polarisations between Ports 2 and 1 and Ports 3 and 1, S21 and S31 respectively, the mode conversion results of the CP element 300 can be found. In addition, the phase difference between the TE10 and TE01 modes can be controlled by tuning the dimensions of the stepped ridge. After an iterative parametric optimization, an amplitude imbalance of less than 1 dB and phase differences from -90° to -77° between TE10 and TE_0i modes, is achieved throughout the frequency band of 10-14.8 GHz. This is reflected in Figure 3(b).

A CP beam can be realized by exciting two orthogonal E fields with equal amplitudes and a phase difference of 90° between them. Base on this concept, a non-ideal CP beam has been formed from the results of Figure 3(b). The axial ratios over frequencies calculated by Equation (1) are shown in Figure 3(c), in which the 3-dB axial ratio (AR) bandwidth is 38.7% (10-14.8 GHz). In Equation (1), D is the phase difference between S21 and S31.

h = \S_2i \² + \S₃₁ \²

In the above equation E_xy refers to the x-direction components of the y-direction electric field - i.e. the projection of Ey (y-direction electric field) on x-direction. Similarly, Sa_b refers to the energy flowing from port b to port a. The characteristic impedance of the radiator of the CP element 300 can be calculated using Equation (3).

The symbols h o and lo herein represent the characteristic impedance of free space, i.e. 377 W, and the wavelength in free space respectively ao is the width of the square radiating aperture as shown in Figure 2(a). Calculated results from Equation (3) show that the characteristic impedances of the radiating aperture vary from 1.17h o to 1.62h o over the frequency band of 10-15 GHz. This is close to the characteristic impedance of free space. A good match between the antenna and the free space will therefore be obtained without any further adjustment after the CP element has been optimized using the aforementioned transmission line analysis method. This is reflected in Figure 4, which shows the simulated reflection coefficients, realized gains and axial ratios of the proposed CP element 300 over frequencies of 9 to 15 GHz. As can be seen, the antipodally ridged CP element has a working bandwidth of 37.4% from 10 GHz to 14.6 GHz. In particular, Figure 4(a) demonstrates a reflection coefficient lower than -14 dB from 10 GHz to 14.6 GHz and Figure 4(b) shows the simulated axial ratio is less than 3 dB throughout the above frequency band. The right hand circularly polarized (RHCP) gain fluctuates from 7.4 dBic to 9.8 dBic within the same frequency band.

To control the ratios of the energies of TEio and TE_0i modes, hi is a key parameter hi is the gap between spaced ridge elements. The transmission line model described with reference to Figure 3 is adopted here to implement a parametric analysis of the amplitude imbalances and phase differences between Port 2 and 3 affected by hi. As shown in Figure 5, | S211 can be smaller, equal or larger than | S311 with variable or variations in hi. This means the distance between the antipodal ridges can be used to control the energy distribution at the output ports hi also affects the phase difference between TE10 and TE01 modes (<S2i - <S3i). To construct the circular polarized antenna, the magnitude imbalance ( | S211 - | S311 ) was designed to approach zero and the phase imbalance or phase difference was designed to approach 90°.

In addition, the CP performance is also related to other dimensions of the ridges (e.g. ridges 116). After implementing the transmission line analysis method to optimise the working bandwidth of the CP element, iterative tuning can be performed to refine or optimise performance of the CP element. For comparison, the present stepped ridge topology was compared with a similar topology having linear shaped ridges. The parametric values of hi, a6, and h2 (see Figure 2) in the design with linear shaped ridges were optimized to be 6.2mm, 19.05mm and 43.5mm respectively to achieve the best performance. The comparison of AR performance of the two designs is shown in Figure 6. As can be seen, the performance of the linear shaped ridge design fluctuated significantly, near sinusoidally over the working band. The present, discrete stepped ridge design, has shorter physical length and better performance than the design with continuous linear ridges.

The antipodal stepped ridges in the multistage cavity may also be changed to other derivatives, such as ridges having greater or fewer steps in their design, while still being optimised according to the transmission line analysis method and iterative parametric optimisation processes described herein. Notably, changing the number of steps in the ridges changes the number of degrees of freedom in design optimisation. In addition, the shape of the ridges, such as discrete, square ridges or rounded ridges, they also be changed depending on space requirements and particular applications.

The CP element may be expanded for use in an array antenna, in which a compact feed network is provided to avail the assembly of higher gains. The array antenna may be uniform or non-uniform. For a uniform array, multiple "H" shaped power divers with wideband characteristics can in some instances be cascaded to assemble all the elements. Such an arrangement works well for microstrip array antennas but is harder to use in the context of waveguide arrays. The radiating aperture of the waveguide element has to be more than half a wavelength to sustain the propagation of the domain mode. In addition, the spacing between adjacent elements should be smaller than one wavelength to suppress grating lobes. An "H" shaped power divider therefore cannot be used to combine four waveguide elements into a 2 x 2 array.

To overcome this problem, higher order mode cavities may be used as compact feeding networks. However, higher order modes narrow the working bandwidth.

Instead, as described herein the problem of feeding the basic 2 x 2 subarray is addressed which makes the array layout easier to design. The compact, wideband feeding network is described herein in the context of feeding a basic 2 x 2 subarray, but can be further expanded to feed larger-scale arrays with high gains and high efficiency. Figure 7 shows the configuration of the proposed compact l-to-4 feeding network, which is also a l-to-4 power divider. The input port is a standard WR-75 waveguide. The contour of the feeding network as an equal or smaller dimension to that of radiating aperture (e.g. Part III - see Figure 2(a)) of the 2x2 array - this promotes high aperture efficiency. The proposed l-to-4 power divider is composed of two orthogonal septa and a stepped cavity. The feeding network combines four CP elements into a 2 x 2 array antenna, comprising a first stage power divider and a second stage power divider. The feeding network may also be extended to larger square array antennas. The feeding network forms a l-to-4 power divider, the design of which is described with reference to eleven cross sections listed and named as surfaces A-K in Figure 7. The first stage power divider is a l-to-2 power divider in the second stage power divider is a 2-to-4 power divider. The l-to-2 power divider is an E-plane power divider realised by surfaces K to G in Figure 7. The 2-to-4 power divider comprises an H-plane power divider (presently two H-plane power dividers) realised by surfaces G to A in Figure 7. The l-to-4 power divider also comprises a first septum 700 in the second septum 702 inserted into the multistage cavity 704 to form the first stage power divider 706 (the E-plane power divider) and the second stage power divider 708 (the H-plane power divider). The septum 700 for the E-plane power divider 706 has a uniform thickness of si. Septum 702 has different thicknesses of si and s2. Both septa 700, 702 intersect with each other orthogonally with their axes of symmetry overlapped with the axis 710 of the stepped cavity 704. The septa 700, 702 extend from the radiating aperture 712 into the stepped cavity 704, with their narrow walls touching the inner surfaces of the stepped cavity 704. The first septum 700 is orthogonally crossed over the second septum 702, extending from the larger open and 714 of the multistage cavity 704, toward a small open end 716. This forms one input port at the smaller open end 716 and for output ports at the larger open end 714.

The design values of the l-to-4 feeding network are listed in Table 2.

Table 2: design values of the compact feeding network. h_AB means the distance between surfaces A and B in Figure 7.

Figure 8 shows the simulated scattering parameters of the compact feeding network. The energy from the input port 716 is equally delivered into the four output ports and end 714 with a maximum insertion loss of 0.16 dB from 10 GHz to 15 GHz, through which the reflection coefficient is less than -15 dB. In addition, the four outports at end 714 are in phase due to the structural symmetry of the feeding network 718.

For experimental purposes, a transition as shown in Figure 1 was designed for connecting together four antipodally ridged CP elements with the compact feeding network to form a 2 x 2 CP array antenna. To improve ease of fabrication, the 2 x 2 CP array antenna was separated into two parts, i.e. Part I and Part II (as shown in Figure 1). Both Part I and Part II were fabricated independently and then connected together using screws, as shown in Figure 9. Part I was printed using an EOS M 290 3D printer and aluminium powder as the material, under direct metal laser sintering (DMLS). Part II was milled from an aluminum block.

Simulated and measured reflection coefficients agree well as shown in Figure 10. The reflection coefficient of the fabricated array antenna, i.e. |Sn| of the fabricated array antenna, is below -10 dB from 9.6 GHz to 15 GHz and below - 15 dB from 9.7 GHz to 15 GHz. Figure 11 shows the comparisons of simulated and measured gains, as well as axial ratios. The measured AR is less than 3 dB within the bandwidth of 9.9-14.9 GHz, throughout which the measured realized gains of the RHCP array antenna fluctuate from 13.2 dBic to 16.3 dBic. Some ripples can be observed from the measured AR curve. These deviations mainly come from the test and fabrication errors. According to Equation (1), the AR values are calculated from two orthogonal vectors which are collected through polarization rotations. The connecting cable may introduce small errors into the amplitude and phase of the orthogonal vectors during the rotation of the mounting platform. Another possible reason is that the present antenna is made using 3D printing technology, for which surface roughness and printing accuracy are not ideal. It is anticipated that improvement 3D printing technology, the differences between the simulated and measured reflection coefficient and other performance parameters will be reduced. These accumulated factors affect the final performance of the CP array antenna. Nevertheless, the measured ARs are still below 3 dB and demonstrate a holistic consistency with the simulations.

Figure 12 shows the simulated and measured co-polarization and cross polarization radiation patterns at 10 GHz, 12.3 GHz and 14.6 GHz. The frequency band of 9.9-14.9 GHz (40.3%) is taken as the working bandwidth for this 2 x 2 CP array antenna, in terms of the overlapped measured bandwidth of axial ratios smaller than 3 dB with that of reflection coefficients lower than -10 dB. Slight differences between simulated and measured results are observable and appeared to be caused by the fabrication error and the measurement system. Despite these slight differences, the fabricated prototype sufficiently approximates the simulated performance that it provides good validation for the present design methodology.

The proposed 2 x 2 CP array antenna can be easily extended to larger-scale arrays while keeping very high efficiency by cascading multiple Ή' shaped power dividers. In addition, the expanded arrays can maintain fixed height - the height being the summation of the heights of the proposed CP array antenna and the extra feeding network. For illustration purposes, and 8 x 8 CP array as shown in Figure 13 demonstrates a simulated total efficiency higher than 78% from 10 GHz to 14.6 or 14.8 GHz (38.7%). Given the difficulty of expanding previously known CP arrays to hire dimensional arrays such as that shown in Figure 13, this further verifies the superiority of the present design methodology, which is based on a 2 x 2 CP array, and the expanded arrays resulting from application of that methodology. When the array scale is expanded from 8 x 8 to 16 x 16, the bandwidth remains the same, although full-wave simulations suggest the total efficiency slightly decreases from 78% to 70%. A comparison between some references and CP arrays produced by the present design methodology is set out in Table 3, the comparison being conducted based on frequency, bandwidth, gain, and efficiency of the respective arrays the total measured efficiency of the present, fabricated 2 x 2 CP array is higher than 80% from 9.9 GHz to 14.8 GHz, although there is a small gain drop at 14.9 GHz. Generally speaking, the present designs are comparatively outstanding in terms of compactness, width of bandwidth, and efficiency.

CP array antenna is formed in accordance with present teachings enable very compact feeding networks to be used. Moreover, compact feeding networks as described herein show wider bandwidth than traditional high order mode cavities and are sufficiently compact to be hidden underneath the radiating apertures, as shown in Figure 1. The CP element and the feeding network can be independently adopted, or can be used together. The proposed CP array antenna with wide bandwidth and high efficiency is an excellent candidate for satellite communications. It will be appreciated that many further modifications and permutations of various aspects of the described embodiments are possible. Accordingly, the described aspects are intended to embrace all such alterations, modifications, and variations that fall within the spirit and scope of the appended claims. Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

The reference in this specification to any prior publication (or information derived from it), or to any matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of endeavour to which this specification relates.

Claims

1. A linear-to-circular polarizer comprising: a cavity; and a pair of stepped ridges disposed in the cavity, the stepped ridges defining a gap therebetween.

2. The polarizer of claim 1, wherein the stepped ridges are disposed on opposite sides of the cavity.

3. The polarizer of claim 2, wherein the stepped ridges are antipodally disposed.

4. The polarizer of any one of claims 1 to 3, further comprising a feeding port and a radiating port, the polarizer polarizing a linearly polarized wave, received from the feeding port, to a circularly polarized wave at the radiating port after propagating through the polarizer.

5. The polarizer of claim 4, wherein the radiating port is a square aperture.

6. The polarizer of claim 4 or 5, wherein the stepped ridges taper towards the radiating port.

7. The polarizer of claim 6, wherein the gap increases towards the radiating port.

8. The polarizer of any one of claims 4 to 7, further comprising a pair of stepped matching blocks disposed in the cavity, the stepped matching blocks being spaced from the stepped ridges and defining a gap therebetween.

9. The polarizer of claim 8, wherein the stepped matching blocks are disposed on opposite sides of the cavity.

10. The polarizer of claim 9, wherein the stepped matching blocks taper towards the radiating port.

11. The polarizer of any one of claims 4 to 10, wherein the stepped matching blocks are laterally offset and each matching block extends into the cavity, starting from the feeding port.

12. A waveguide circularly polarized (CP) element antenna comprising: d) a feeding waveport; e) a linear-to-circular polarizer according to any one of claims 1 to 11; and f) a radiator for radiating the circularly polarized electromagnetic wave.

13. A feeding network for combining four CP elements into a square array antenna, comprising a first stage power divider and a second stage power divider.

14. The feeding network of claim 13, wherein the first stage power divider is a l-to-2 power divider, and the second stage power divider is a 2- to-4 power divider.

15. The feeding network of claim 14, wherein the l-to-2 power divider is an E-plane power divider and the 2-to-4 power divider is composed of two H-plane power dividers.

16. The feeding network of any one of claims 13 to 15, further comprising a first septum and a second septum inserted into a multiple stage cavity to form the first stage power divider and the second stage power divider.

17. The feeding network of claim 16, wherein the multiple stage cavity has a smaller open end and a larger open end, the larger open end being larger than the small open end.

18. The feeding network of claim 16 or 17, wherein the first septum is orthogonally crossed over the second septum, extending from the larger open end toward the smaller open end to form one input port at the smaller open end and four output ports at the larger open end, both septa with different dimensions touching lateral walls of the multiple stage cavity.

19. The feeding network of claim 18, wherein the first septum has a consistent thickness and the second septum has a first thickness and a region of a second thickness, thicker than the first thickness.

20. An array antenna assembly comprising: four waveguide CP element antennas, each waveguide CP element antenna being a waveguide antenna according to claim 12; and a feeding network according to claim 18 or 19, the output ports being connected to the feeding ports of all four waveguide CP element antennas by a waveguide transition.