Dual-polarized antenna

[0002] This application claims priority under 35 U.S.C. §§119 and/or 365 to 9903920-8 filed in Sweden on Oct. 29, 1999; the entire content of which is hereby incorporated by reference.

[0003] This invention relates to the technical field of antennas, and more specifically to dual-polarised antennas.

[0004] Historically, the most common diversity technique, especially in mobile communications, has been space diversity. In this diversity technique two or more antennas are placed apart from each other, separated by a distance that is a function of the wavelength the antennas should receive or transmit. On uplink the incoming signals received at each antenna are combined in an optimum way so as to maximise the quality of the resulting signal.

[0005] Owing to economical as well as aesthetical reasons, base stations are nowadays increasingly equipped with a single dual-polarised antenna, or one such antenna for each sector or frequency band. Such an antenna must provide two at least mainly orthogonal polarisation directions in order to enable the use of diversity techniques. This is possible, as orthogonally polarised waves are essentially uncorrelated in a multipath environment.

[0006] Good transmission and reception characteristics are obviously important in antennas. For dual-polarised antennas, these characteristics are among other things affected by isolation between the antenna's feed ports, and by the farfield pattern characteristics. Good isolation between feed ports is much more difficult to obtain in a dual-polarised antenna than in a pair of space diversity antennas as the latter are located a distance apart from each other.

[0007] The isolation problem is easier to describe if an antenna is considered as analogous to a four-port. Two of the four ports represent actual transmission line feed ports; one for each of the two desired orthogonal polarisations. The other two ports are virtual ports representing, on transmit, radiated power in each of the two orthogonal polarisations, integrated over an arbitrary sphere enclosing the antenna. Thus, on transmit, the antenna has two input ports (feed ports) and two output ports (radiated power). Similarly on receive the antenna has two input ports (received power) and two output ports (feed ports). Note the the same feed ports are used for both transmission and reception. The antenna, and its four-port representation, is reciprocal.

[0008] The scattering parameters of a four-port are often represented by the S-matrix, a four-by-four matrix. The S-matrix for an ideal dual-polarised antenna has four non-zero values all of unit magnitude. These values represent the forward and backward coupling between corresponding input and output ports.

[0009] Isolation between ports is never perfect in practice. This leads to a certain degree of mutual coupling. From a system point of view, the study of mutual coupling effects can be limited to two categories: isolation between the feed ports and coupling between feed port and undesired output port. Isolation between feed ports is primarily a problem on tranmit while coupling between feed ports and undesired output ports is primarily a problem on receive.

[0010] Isolation between feed ports is of primary importance in the transmit direction, i.e. in the downlink band for mobile telecommunication. In a base station antenna, transmitted power is many orders of magnitude greater than received power. It is therefore important to stop transmitted signals from leaking into the received signal paths. This is achieved with filters, and the worse the mutual coupling between feed ports the worse the leakage and the better the filters have to be. Better filters, which give better suppression, are as a rule more expensive and unwieldy. On the other hand, reduced mutual coupling between feed ports enables the use of simpler and less expensive filters.

[0011] Mutual coupling between feed ports can also cause problems when the transmitted signal from one port travels “backwards” via the other transmit branch. This may give rise to intermodulation effects and spurious radiation, both in the uplink and downlink frequency bands.

[0012] Coupling between a feed port and undesirable output port will result in polarisation impurities. On receive the impurities may increase the correlation between the received signals, which in turn diminishes the diversity gain potential. The similarity between the angular power distribution of the radiation patterns for the two polarisations will also affect the attainable diversity gain. In general, the more alike the radiation patterns of the two polarisations are, the better the antenna will be from a polarisation diversity point of view. Furthermore, equal patterns for both polarisations are important on both uplink and downlink in order to cover the same angular and radial region.

[0013] There are a number of factors that adversely affect the polarisation purity and the mutual coupling of an antenna. The finite size of the antenna will give rise to both of the above owing to edge and corner diffraction. The antenna may consist of one, two or more primary radiators (antenna elements). Dual-polarised primary radiators also generate mutual coupling and polarisation impurities. This is owing to asymmetries in both the radiating elements and the feed network. Mutual coupling can also be caused by the proximity of the element feed points.

[0014] Several attempts have been made to overcome these problems with dual-polarised antenna elements. Some of these attempts will be described below.

[0015] A common type of dual-polarised antenna element is the aperture-coupled microstrip patch antenna. By feeding a patch using orthogonal slots (apertures) the antenna is made to simultaneously radiate and/or receive two orthogonally polarised waves. Similar characteristics can be achieved by using probe-fed patches, but the aperture-coupled patch is superior to the probe-fed patch from a bandwidth, passive intermodulation and manufacturing point of view and is the dominant type in use today for communication applications.

[0016] U.S. Pat. No. 4,903,033 presents a dual-polarised antenna that involves an air bridge to accomplish a symmetric feed arrangement in both of the polarisation branches, while Sanford, J. R. and Tengs, A.: “A Two Substrate Dual-Polarised Aperture-Coupled Patch”, Proceedings 1996 IEEE Antennas Propagation Society Symposium, present an antenna where symmetry is achieved by placing the feed networks on opposite sides of a dielectric substrate, with the feed slots de-embedded in the centre of the substrate. Both solutions involve feed network layouts that make the dual-polarised patch element more complex than a single polarised element. Furthermore, neither of the two solutions is truly symmetric. Mutual coupling, primarily between the two crossing feed arms, will introduce asymmetries in the excitation current. These asymmetries can only be compensated for at one or a few frequencies, if at all.

[0017] U.S. Pat. No. 5,045,862 presents a microstrip array arrangement useful for reception with a high degree of symmetry. This single layer arrangement consists of interconnected etched square patch elements and filters. These elements and filters make up a square periodic grid, in two orthogonal directions. Each patch is connected to its four neighbouring patches by symmetric feed lines extending from the centre of the four sides of the patch element. While this arrangement is truly symmetric, it is not useful for communication applications for a number of reasons. For example, radiation may occur from feed lines, filters, and matching stubs when using one side of the dielectric layer, and the coupling between polarisations is not suppressed by the feed arrangement.

[0018] WO 98/49741 proposes a compact and simple feed solution, see FIG. 1. A single layer feed-network design is utilised with a combination of a symmetrical and an asymmetrical feed arrangement. This solution can only be used to obtain a two-port antenna element with good isolation properties at one or a few frequencies. Nor does the design allow for simultaneous symmetrical slot feed for both polarisation ports.

[0019] For the reasons stated above it is desired to find a dual-polarised antenna element with improved transmission and reception characteristics.

[0020] The present invention aims to solve the problem of how to improve transmission and reception characteristics in dual-polarised antennas.

[0021] One object of the present invention is to provide a dual-polarised antenna element for which the transmission and reception characteristics are improved owing mainly to the layout of the feed network.

[0022] Another object of the present invention is to provide a layout of a feed network for a dual-polarised antenna in order to reduce or cancel undesired effects from one part of the network to another and vice versa.

[0023] Yet another object is to provide a method for feeding current to two orthogonal polarisations in a dual-polarised antenna element.

[0024] Still another object is to provide a method for obtaining a dual-polarised antenna element of the above-mentioned kind.

[0025] According to the present invention, there is provided a dual-polarised antenna where the effect of the signals feeding one polarisation, on the other polarisation, is cancelled owing to the layout of the lines through which the signals are fed.

[0026] According to the present invention, there is provided a method for feeding current to two orthogonal polarisations in a dual-polarised antenna element so that the effect of the current on the other polarisation is cancelled owing to the layout of the lines through which the signals are fed.

[0027] According to the present invention, there is also provided a method for obtaining a dual-polarised antenna element of the above-mentioned kind.

[0028] The apparatus according to the invention is defined in claim 1.

[0029] Preferred embodiments of the apparatus according to the invention are defined in claims 2-24.

[0030] The first method according to the invention is defined in claim 25.

[0031] The second method according to the invention is defined in claim 26.

[0032] Preferred embodiments of the second method according to the invention are defined in claims 27-35.

[0033] An advantage with the present solution to the problem is that the port isolation in the dual-polarised antenna is improved.

[0034] Another advantage with the present solution to the problem is that the similarity of the dual-polarised antenna's patterns of the two polarisations is increased.

[0035] The invention will now be described in more detail with the aid of the description of the embodiment and with reference to the drawings.

[0036] FIG. 1 shows an example of an existing feed network found in prior art.

[0037] FIG. 2 shows the parts of a slot coupled microstrip patch element.

[0038] FIG. 3 shows a dual-polarised microstrip antenna element according to the invention.

[0039] FIG. 4 shows another embodiment of a dual-polarised microstrip antenna element according to the invention.

[0040] FIG. 5 shows yet another embodiment of a dual-polarised microstrip antenna element according to the invention.

[0041] FIG. 6 shows one more embodiment of a dual-polarised microstrip antenna element according to the invention.

[0042] FIG. 7 illustrates an advantage with an embodiment of the dual-polarised microstrip antenna element according to the invention.

[0043] FIG. 1 shows an example of a prior art dual-polarised microstrip antenna (WO 98/49741). An antenna element 10 comprises a patch 11 and two orthogonal slots 12a, 12bas feed structures. Two ports 13a, 13b—one for each polarisation—are the sources for the feed network 14. Port 13ais connected to network part 14athat bifurcates into branches 14a1 and 14a2. Each of these branches 14a1, 14a2 cross the vertical slot 12a, one on each side of slot 12b. Port 13bon the other hand is connected to another network part 14b1. This second network part 14b1 intersects slot 12b.

[0044] As can be seen from FIG. 1, slot 12bis not fed symmetrically. The design does not allow for simultaneous symmetrical slot feed for both polarisation ports. This leads to polarisation impurities.

[0045] One way of improving port isolation is to arrange and exploit symmetries in the feed network. The resulting current symmetries will also generate similar patterns for the two orthogonal polarisations.

[0046] A first concern when designing the feed network is to mitigate the coupling effects. Placing any two transmission lines as far from each other as possible is a way of doing this. Transmission line losses must however also be taken into account. In addition, discontinuities, e.g. bends, in the transmission lines should preferably be avoided. When such discontinuities are unavoidable, the layout should be chosen so that as little spurious radiation as possible should be radiated by them.

[0047] One way of eliminating the mutual coupling effects is to choose the feed network layout so the mutual coupling effects of individual coupling contributions cancel each other when summed over all components. This is achieved by having each feed line with a certain current (or voltage) matched with an identical mirror-imaged second line with the identical current and amplitude as the first line. This latter current should either be in-phase or 180 degrees out-of-phase, depending on the layout, when compared to the first current.

[0048] Pattern similarity is related to the symmetry of the antenna element. Preferably, all parts of the element should exhibit symmetry properties. This includes both patch and slot symmetries as radiation fields will derive from both the patch currents and the slot fields.

[0049] The pattern similarities also depend on the geometry of the patch. Patches with at least two orthogonal symmetry planes such as circular or square can be used since they reduce mutual coupling effects. Circular patches have the advantage of being less sensitive to for instance manufacturing tolerances with respect to rotational installation.

[0050] FIG. 2 shows the parts of a single polarised slot coupled microstrip patch element. This is intended to facilitate the comprehension of the following figures. As in FIG. 1, a patch is indicated by 11, a feed structure, in this case a slot, by 12 and the feed network by 14. As can be seen these three parts are located in different planes and the slot 12 is an aperture in a ground plane 9.

[0051] Where it in the description below says that any part of the feed network 14 enters the patch 11, this is only a way to facilitate the reading of the description. It should always be understood that the feed network 14 is in a plane of its own, separate from the plane belonging to the patch 11. Strictly speaking, where, in this context, it says “patch 11” it should usually be read “projection of said patch 11” and so on.

[0052] FIG. 3 shows a possible embodiment of a dual-polarised microstrip antenna element according to the invention, a square slot-coupled microstrip patch antenna. In this figure, an antenna element 10 comprises a square patch 11. The patch 11 in this figure is planar, but non-planar patches can also be used. The antenna element also comprises a number of slots 12a-12das feed structures. The staple-shaped slots 12a-12dare symmetrically arranged, one in the centre of each of the patch's 11 sides. At least parts of the legs of a staple usually protrude beyond the projection of the patch 11, while the overlying bar remains within the contour of the projection of the patch 11. The signals for the different polarisations are fed into the feed network 14 at the ports 13a, 13b. Each port 13a, 13bis connected to two opposing slots 12a-12d. In this figure, port 13afeeds slots 12band 12d, while port 13bfeeds slots 12aand 12c. As each port 13a, 13bfeeds slots 12b, 12dand 12a, 12c, respectively, the feed network 14 divides into two branches from each port 13a, 13b. Port 13ais connected to branch 14a1 and 14a2, while port 13bis connected to branch 14b1 and 14b2. The branches 14a1, 14a2, 14b1, 14b2 all enter a projection of the patch close to a corner and leads further in on a diagonal or near-diagonal, in order to be equally distant from the nearest slots 12a-12das the other branch 14a1, 14a2, 14b1, 14b2 belonging to the same network part 14a, 14b. After having entered the patch 11, the branches 14a1, 14a2, 14b1, 14b2 cross the center of the slots 12a-12dwhile conforming to the design rule of maintaining symmetry.

[0053] The dashed parts of the feed network 14 symbolise a change in phase of the signal 180 degrees effectively, with regard to the signal in the other branch of the same network part 14a, 14b. Here, and analogously elsewhere, it means that the phase also could be shifted for instance 540 or −180 degrees, as 0 and 360 degrees are equivalent. That is, the electric length of the dashed part is 180 degrees. For example, the signal running in branch 14b1 will be phase shifted 180 degrees compared to the dashed part of said branch 14b1. After said dashed part the signal will at least essentially have the same magnitude as the signal in the other branch 14b2, but the phases of the signals will be 180 degrees apart. The coupling between the slot 12band the two branches 14b1 and 14b2 respectively will then have the same magnitude but will be 180 degrees out of phase. This results in a cancellation of the coupling effects, as summation of the two coupling components will cancel each other in slot 12b. This works analogously for the coupling components with slots 12dfed from the port 13b.

[0054] The above-mentioned 180 degree phase shift could, as is known in the art, be attained by line length differences or Shiffman phase shifters. The phase shift can be 180 degrees plus an integer times 360 degrees. Possible differences in signal strength in the two branches owing to transmission line losses can be compensated for if desired. Furthermore, the slot geometry can be chosen to conform with the current distribution on the patch. By using a shaped slot aperture and a non-uniform slot width, with the slot geometry matched to the current distribution on the patch, the overall electric properties of the patch and the slot as an entity are optimised for maximum performance.

[0055] FIG. 4 shows another embodiment of a dual-polarised microstrip antenna element according to the invention. In this figure, the ports 13a, 13b, and the two branches 14a1, 14a2 remain the same as in FIG. 3. The differences are the shape of the patch 11, the shape of the slots 12a-12d, and the layout of the branches 14b1, 14b2. The patch 11 is circular and slots 12a-12dare shaped like bent staples; the legs of the staple are straight while the overlying bar is bent. Said legs are positioned mainly outside the patch 11. The two branches 14b1, 14b2 are, conforming to the layout rules, arranged outside the patch until they cross the middle of the slots 12a, 12cfrom the outside leading in. The cancellation of coupling effects works as described for FIG. 3.

[0056] FIG. 5 shows yet another embodiment of a dual-polarised microstrip antenna element according to the invention. As before, 10 indicates the antenna, 11 the patch, 12 the slots, 13a, 13bthe ports, and 14 the feed network. The difference between this figure and FIG. 3 is the layout of the network part 14a, leading from port 13a. From said port 13a, positioned straight outside the centre of slot 12d, a branch 14a3 runs straight across said slot 12d. The other two branches 14a1, 14a2 run symmetrical with regard to a line intersecting the middle of slots 12band 12d. Said branches 14a1, 14a2 surround slot 12dby running essentially parallel to its upper bar for a while, then turning to enter the patch 11 on the closest diagonals, intersecting in the middle of the patch 11, after which a single line crosses the opposite slot 12bfrom the inside out.

[0057] The dashed part of branches 14a1 and 14a2 shifts (relative to 14a3) the phase of the signal 360 degrees—or any integer times that. This will cause the signals in the branches 14a3 and 14a1, 14a2 to be in phase when crossing the slots 12band 12d.

[0058] FIG. 6 shows one more embodiment of a dual-polarised microstrip antenna element according to the invention. As before, 10 indicates the antenna, 11 the patch, 13a, 13bthe ports, and 14 the feed network. The differences between this figure and FIG. 3 are that there are no slots in the antenna 10 in FIG. 6, and that the layout of the feed network 14 is slightly different. Instead of slots, the feed structures comprise probes 15, for instance galvanic or capacitive, feeding through a ground plane (not shown). The feed network 14 in FIG. 6 is laid out so that the branches 14a1 and 14a2 end inside the patch 11. The other two branches 14b1 and 14b2 are shorter as well. All the branches 14a1, 14a2, 14b1, 14b2 end the same distance from the edges, in the centre of the sides of the patch 11, where they are connected to the probes 15.

[0059] FIG. 7 illustrates the cancelling of coupling phenomena for the embodiment described in FIG. 3. New in this figure are six dotted ellipses 20a-20frepresenting regions of coupling fields and a number of arrows indicating the phase of currents and coupling fields. Assume that power is fed into the upper port 13b. Both branches 14b1 and 14b2 cause coupling components in the slot 12b, indicated by ellipses 20aand 20b. As part of branch 14b1 is designed to shift the phase of the signal 180 degrees, the coupling components cancel each other in the slot, as they have the same magnitude while being 180 degrees out of phase. Furthermore, the coupling between slots 12aand 12cand network part 14acancel for the same reason at port 13a, indicated by ellipses 20cand 20d. Both branches 14b1, 14b2 cause coupling components at branch 14a1 indicated by ellipses 20eand 20f. Again, the coupling components cancel since the signals in branches 14b1 and 14b2 are 180 degrees out of phase. By reciprocity, the same cancelling effects will occur for signals from the other port, 13aat port 13b.

[0060] The cancellation effects in a probe-fed antenna element, FIG. 6, can be explained in a similar way as in FIG. 7 since the feed network 14 exhibit the same symmetry properties.

[0061] There is also provided a way (or method) of feeding two orthogonal polarisations. A benefit of this is that undesired coupling effects between polarisations are cancelled. This is achieved by using any of the embodiments of the apparatus according to the invention or any other combination of antenna parts that is essentially equivalent.

[0062] A possible field of application for the apparatus according to the invention is in an array antenna. This kind of antenna comprises many antenna elements, of which some or all can be of a kind described above.

[0063] The arrangement according to the invention is not necessarily limited to the way it was described or presented in the drawings, as they are intended to give an understanding of the general idea. The shape of the slots 12 could be different as long as the general idea is conformed with. Similarly, the layout of the feed network 14 is not restricted to the exact designs given above, but could vary to a certain extent, while retaining the fundamental symmetry characteristics described above. The thickness of the lines in the feed network 14 are not necessarily drawn to scale; but are drawn to facilitate comprehension.