Field of the Invention
The invention concerns microstrip antenna devices.
Prior Art
Many antenna structures have already been described in this field. The simplest microstrip radiating structure includes a dielectric layer carrying on one side a conductive patch of a chosen shape and on the other side, a conductive plane called a ground plane. To obtain an antenna, it is necessary to define the mode of feeding this structure with ultra-high frequency energy.
The idea of providing a stack of superposed patches has already been described in LONG & WALTON's article "A dual frequency stacked circular disc antenna", IEEE Transactions on Antennas and Propagation, Vol. AF 27, No. 2, March 1979. Other proposals have since been formulated.
As regards the feeding of superposed two-patch antennas, two very different cases have to be distinguished from the point of view of operation, according to whether the feeding is effected at the upper patch or the lower patch (the one nearer the ground plane).
In the case where the feeding is obtained at the lower patch, this is most frequently a connection at the circumference of this patch. Moreover, provision is systematically made for an upper patch of a larger size than that of the lower patch (see in particular the article of TULINTSEFF, ALI & KONG, "Input impedance of a probe-fed stacked circular microstrip antenna") IEEE Transactions on Antennas and Propagation, Vol. 39 No. 3, March 1991).
The expert will know that the perfection of superposed patch antennas is particularly intricate. Attempts have been made to model their properties. By way of example, we will mention COCK & CHRISTODOULOU's "Design of a two-layer, capacitively coupled, microstrip patch antenna element for broad band applications", IEEE Symposium on antenna propagation, 1987. In spite of these attempts, it is still extremely difficult to predict by modelling, and to understand the behaviour of microstrip structures comprising two or more superposed patches.
The Applicants' assignees have set themselves the problem of obtaining a conformable antenna with electronic scanning, intended for the communication system with movable objects such as aircraft (the system called SATCOM).
This system is provided for operating with the group of geostationary satellites managed by the INMARSAT organisation. At least as far as the applications to aircraft are concerned, the proposed telecommunications service is governed by an international standard called ARINC 741.
Technically, one is concerned with setting up an antenna capable of operating, on the one hand, in the transmitting mode and, on the other hand, in the receiving mode, in two very close bands, that is to say, one a little higher than 1.5 gigahertz for receiving and another a little higher than 1.6 gigahertz for transmitting.
The electronic scanning function is necessary for this antenna because of the movement of the movable carrier which is here assumed to be an aircraft. It is also necessary to choose between a roof antenna or two lateral antennas. In the case of two lateral antennas, the above mentioned ARINC Standard has defined two official acceptable shapes defining the volume into which the planned antenna has to be fitted.
The antenna must also be conformable, that is to say, be capable of adapting to the exact wall-shape of the movable carrier. It must, moreover, be thin so as to minimize the aerodynamic drag and of course, be designed so as to comply with the required mechanical characteristics required for the structure of the aircraft.
During the research they have undertaken, the Applicants' assignees have found that it was possible to design a microstrip antenna going virtually against the solutions accepted so far by the experts.
Object of the Invention
It is therefore an object of the present invention to provide an antenna element that is fundamentally different from those known so far.
It is a further object of the invention to provide an antenna element of the type comprising a first dielectric layer including on one side a ground plane and on the other a first conductive patch of a chosen shape, a second dielectric layer which surmounts the first layer on the side of the first patch and which supports on the other side opposite the first patch a second conductive patch of a chosen shape, a third dielectric layer surmounting the second, and also means for feeding ultra-high frequencies to one of the conductive patches.
Summary of the Invention
According to the invention, the second patch is of a smaller size than that of the first patch and the electrical connection to this first patch is from below at at least one chosen point situated between its center and its circumference.
With this structure, it has proved possible to construct an operational antenna, subject to choosing the position of the connection point in question according to the respective sizes of the first and second patches, and according to the dielectric characteristics of the first and second dielectric layers, as well as those of the third dielectric layer which preferably has dielectric constants that are distinctly higher than those of the two others.
The first patch may be connected to a lead-in through the ground plane joining a feeding circuit implanted in a dielectric substrate of a three-plate-type structure. More particularly, the three-plate structure includes a substrate layer implanted between the above mentioned ground plane and a bottom ground plane; between the two ground planes, provision is made for conductive lead-ins defining a peripheral shield for the feeder part of the antenna element. Preferably, provision is made for a Wilkinson divider capable of feeding the lower patch at two points which together with its centre, form a substantially right-angled isosceles triangle, while the respective signals brought to these two points are in quadrature. The Wilkinson divider is implanted at an intermediate level of the substrate layer in accordance with the three-plate structure. This intermediate level serves in practice as the feeding distribution level between a central connector for the antenna as a whole and the various antenna elements which, in the application as an antenna array will constitute the antenna as a whole.
In an advantageous embodiment, the two patches have a generally circular shape and these two patches are substantially coaxial, that is to say, they are situated on the same perpendicular to the planes of the dielectric layers.
Brief Description of the Drawings
Other objects and advantages of the invention will become apparent on examining the detailed description given below and the attached drawings wherein:
FIG. 1 is a general schematic diagram of an antenna element in an exploded perspective;
FIG. 2 is a broken partly sectioned view of an antenna element;
FIG. 3 is a (superposed) detailed part view of the connection of the lower patch to its feeding means by a Wilkinson divider;
FIG. 4 is a view from below of the twenty four Wilkinson dividers, for a 24 element antenna, interconnected to the central connector;
FIG. 5 is a top view of twenty four lower patches corresponding precisely to FIG. 4; and
FIG. 6 is a diagram showing the reflection coefficient of the antenna in relation to the frequency.
Detailed Discussion of Preferred Embodiments
The expert will know that shape is important in microstrip devices. Moreover, the drawings are in essence of a definitive nature. They may therefore be incorporated in the description not only to render the latter more readily understood but also to contribute to the definition of the invention if required.
In FIGS. 1 and 2, the reference PMO designates a bottom ground plane which may be fitted by means of an insulating adhesive, on a sheet to be incorporated in the wall of the aircraft. This bottom ground plane is surmounted by two dielectric layers SDB and SDH (low and high respectively). The layer SDH is in turn surmounted by another ground plane PM1. The whole forms a three-plate structure with appropriate metallisations engraved between the layers SDB and SDH or more precisely, on one of these layers.
Fundamentally, these metallisations include a feeder line L which is subsequently subdivided in the manner of a Wilkinson divider, which is schematically outlined in FIG. 1 but is more clearly seen in FIGS. 3 and 4. This divider comprises two branches DL1 and DL2 which first diverge, to rejoin each other in a region where they are-connected to a resistor RLL implanted in the thickness of the layer SDB, but without rejoining the bottom ground plane PMO. Subsequently, the two branches DL1 and DL2 again diverge, to rejoin the respective connection points EL1 and EL2.
These points EL1 and EL2 are connected via lead-ins TR1 and TR2 (not connected to the ground plane PM1) to connection points FR1 and FR2 provided on the lower patch or control patch, P1 engraved on the top face of a dielectric layer D1 placed above the ground plane PM1.
As may be seen in FIGS. 3 and 4, the end portions of the engravings DL1 and DL2 have different lengths, so that electromagnetically, the signals available at the level of points FR1 are substantially in quadrature with each other. The connection points FR1 and FR2 of the patch P1 are situated on respective radii which are substantially at right angles to each other.
The distances d1 and d2 of these points from the centre of the patch P1 are in principle equal. The choice of these distances will be reverted to below. But it is possible to indicate forthwith that these distances d1 and d2 are in principle comprised between 50% and 100% of the radius of the patch P1 (designated DP1/2 in FIG. 3).
Above the patch P1, a second dielectric layer D2 is provided having the same dielectric constant as the layer D1 but having a greater thickness, as may be seen in FIG. 2. In the upper portion, the layer D2 receives by engraving a second conductive patch P2 (a coupled patch) which is generally circular and coaxial with the patch P1, but has a shorter diameter than that of the patch P1.
The antenna element is completed by an additional dielectric layer DR forming a radome and having in principle a dielectric constant that is distinctly higher than that of the layers D1 and D2.
In FIGS. 2 and 4, it will moreover be seen that the line L continues as far as a passage via a metallised hole to a generally coaxial-type ultra-high frequency connector CCH situated behind the metallic sheet subjacent to the bottom ground plane PMO.
Moreover, comparing FIGS. 2 and 4, it will be seen that this connector is provided for each contact stud with a horseshoe-shaped peripheral shield passing through the whole of the dielectric layer SDB. This shield could be defined by a continuous conductive layer. The Applicant has found that it was sufficient to make provision for a certain number of traversing studs surrounding the location of the lead-in CCH, with an interspacing between these studs which remains sufficiently shorter than the wave length of the ultra-high frequency signals processed.
Similarly, the peripheral studs such as BP11, BP12 and BP13 define a shield for the feeding of the antenna element in question, relative to the neighbouring antenna elements and with respect to the outside.
It will, however, be noted that above the ground plane PM1, no provision is made for any insulation of the antenna element relative to its neighbouring elements.
FIG. 5 shows how 24 antenna elements may be disposed to form a conformable antenna with electronic scanning, satisfying the conditions of the problem posed. As has already been indicated, these antenna elements are connected to a general connector with (at least) 24 pins. Up the line from this connector, provision is made for an individual reciprocal phase shift treatment for each antenna element by means of controllable phase shifters DPH schematically outlined in FIG. 2.
The main parameters affecting such an antenna are:
the height and the dielectric constant of the three layers DR, D2 and D1;
the diameters of the patches P1 and P2 and
the radii d=d1=d2 of the two feeding points of the bottom patch P1.
The problem posed in the particular intended application is to obtain a dual behaviour from the unitary antenna element (FIG. 6) namely:
a) a dual frequency behaviour including a very good adaptation (better than -20 decibels) on two frequencies F1 and F2;
b) a broad band characteristic ensuring at least an adaptation of -10 decibels between the frequencies Fr and F4 containing the frequency interval of F1 and F2.
The applicant has observed that provided the frequencies F1 and F2 are not too remote from each other and, seeing that the parameters of the heights and dielectric constants of the above mentioned three layers are fixed, there exists in practice only one solution in terms of the radii of the two patches and of the feeding radius of the patch P1 which would make it possible to satisfy the conditions set out above.
Any modification of one of the parameters has the effect that it becomes very difficult to rediscover a situation capable of satisfying the conditions.
Although the phenomena in question have not yet been completely understood, it seems that in the general case, everything is taking place as though only one of the two patches P1 and P2 resonates at the operating frequency. On the other hand, there exists a very small domain in the parameters for the definition of the antenna, wherein the two patches are interacting while showing a typical dual frequency behaviour as desired. It is still necessary to search for the optimum point of this dual frequency behaviour to respond to the desired operating conditions for the antenna, such as those set out above.
In particular, it has been shown that in practice it is very difficult to cause the antenna element to function without adding thereto a top radome layer DR.
The Applicant has thus been able to obtain antennas responding to the following parameters:
thickness of the layer DR: 1.5 to 2.5 mm;
relative dielectric constant of the layer DR: from 4 to 5, and in a preferred embodiment, on the order of 4;
thickness of the layer D2: approximately 4.8 mm;
thickness of the layer D1: approximately 1.6 mm;
relative dielectric constants of the layers D1 and D2 as well as SDB and SDH: on the order of 2;
diameter of the patch P1: approximately 70 mm;
diameter of the patch P2: approximately 60 mm;
radius of the feeding points FR1 and FR2: from 0.5 to 0.7 times the radius of the patch P1.
Such antennas can satisfy the stipulated conditions for the SATCOM operating band, that is to say:
a reflection coefficient better than -20 dB at the central receiving frequency (1.545 GHz);
a reflection coefficient better than -20 dB at the central transmitting frequency (1.645 GHz);
where, the frequency of 1.545 GHz has a wavelength of 194 mm and the frequency of 1.645 GHz has a wavelength of 182 mm and the diameter of the first conductive patch is less than one half either of the wavelengths and, in a preferred embodiment, the diameter is between 36% and 38% of the wavelength;
a band pass characteristic at a level better than -10 dB between 1.53 and 1.66 GHz.
There will now be discussed the setting up of an array of antenna elements such as illustrated in FIGS. 4 and 5.
First of all, it has been indicated above that each bottom patch is fed at two points situated on respective radii which are substantially perpendicular to each other.
It has appeared worthwhile to distribute the two connection points in a suitable way and this in a different manner for the 24 antenna elements illustrated. The Applicants have found that this makes it possible to reduce the ellipticity (elliptical eccentricity) of the antenna, taking into account that it operates in the circular polarisation mode and with electronic scanning. For this purpose, it is possible either to distribute the connection points substantially at random or to search experimentally for an optimum configuration from the point of view of this ellipticity (for example, as in FIG. 5).
The thus obtained antenna array with electronic scanning has proved capable of operating with loss of aim (scatter) angles of up to 60.degree., with sufficiently low secondary lobe levels, and with a gain of at least 12 decibels as compared with an isotropic antenna.
A good compromise between the loss of gain and the secondary lobe level has been obtained by applying a slightly amplitude-weighted law of illumination. This may be a Taylor law of the circular 20 decibel type, these indications being comprehensible to the expert.
The phase shifters associated with each of the antenna elements may be integrated in the beam steering unit (or BSU) accommodated inside the aircraft.
Advantageously, line phase shifters are used that are switched by PIN diodes controlled by four bit binary words, whereby a resolution of 22.5.degree. is obtained.
The distributor integrated in the phase shifter block ensures the amplitude weighting according to the above mentioned law.
In the particular intended application, the antenna must operate simultaneously in the transmitting and receiving modes at relatively close frequencies. As regards the calibration of the electronic scanning phase shifters, it is necessary to place the array in phase or to "phase" it over a band of approximately 8%.
Rather than calculate the phase code at the central frequency of the band, the Applicants have found that it was preferable to take into account the use of the two distinct frequencies, as well as the quantification and the nature of the phase shifters (switched lines). For this purpose, they use the calibration procedure described below.
Let an element Ai be taken of a conformable, hence non-planar, antenna, with coordinates (at the centre) Xi, Yi, Zi. When it is desired to displace the main beam into the direction U, V at the frequency f, it is necessary to apply to this antenna element Ai a theoretical phase shift DPi which is a function (known to the expert) of f, U and V:
In practice, a calibration table TC (n, F) is used, where n is an integer (or another discrete variable) representing the required state of the phase shifter, with 0.ltoreq.n.ltoreq.N, while one also limits oneself to discrete values for the frequency F. This is written as:
In the intended example, 101 frequency points are taken in the 1.53-1.66 GHz band; and N=15, with n defined in 4 bits. This method only "phases" the array correctly for a single frequency. Now the antenna essentially has a dual frequency behaviour.
The Applicant has then established a "distance" between the theoretical phase and the tabulated phase for the two frequencies f1 and f2, in particular in the form of:
where .vertline. designates the absolute value (modulus).
The calibration then lies in looking in respect of each aiming direction and each antenna element a priori for the value n which minimises this function DDi.
The actuation of the phase shifters is effected accordingly. This calibration can, of course, be stored.
The present invention is not necessarily limited to the embodiment described, nor to the application intended. The antenna element may itself be used for other applications provided the new structure is retained. The combination of a microstrip element and a three-plate feeding arrangement in the same dielectric stack also merits consideration.
The polarisation may be other than the circular polarisation of the embodiment described.
Another particular feature of the invention is that it can avoid, as far as the layers D1 and D2 are concerned, recourse to dielectrics with a low constant, or porous dielectrics or even those constituted by a gas.