Patent Application
20050237238
The invention belongs basically to the field of small-sized radio antennas. Especially the invention is related to utilizing a space-filling curve in the design of an antenna for a portable communications device.
The portable communications devices of modern telecommunications systems need antennas that should fulfil a number of requirements, some of which appear to be mutually contradictory. The antenna should be small, light and easy to manufacture in large-scale mass production at low cost. The antenna should have resonant frequencies in multiple frequency ranges, which in cellular communications systems are up to 1000 MHz apart from each other, and in FM radio reception can be as low as below 100 MHz. The input impedance of the antenna should match the impedance of an antenna port of a transceiver or receiver over a relatively wide frequency band. Losses in the antenna, caused by conduction losses in the conductive parts of the antenna and dielectric losses in the supporting and surrounding materials, should be as low as possible.
Especially the requirement for a small size causes difficulties. In general, the smaller the antenna is made, the narrower its impedance bandwidth becomes. The miniaturization requirements concern not only the radiating antenna part; also the ground plane related to the antenna structure should be as small as possible.
Interesting developments in this field have been introduced in the form of fractal antennas. A fractal is a self-similar structure, which means that a small part of the structure is a scaled-down copy of the original structure. A fractal antenna is one where a radiating antenna element has the shape of a fractal curve. The self-similarity of the structure often leads to multifrequency operation, because at a higher frequency and thus a smaller wavelength a smaller part of the antenna replicates the resonant characteristics of the whole antenna at a lower frequency. A fractal curve is also relatively long compared to the overall two-dimensional area it occupies. This is advantageous, because the end-to-end length of a line-shaped antenna radiator must be at least one quarter of the wavelength at the desired resonant frequency. It is relatively easy to make a small-sized antenna structure by using a tightly meandering fractal curve as the radiating part.
Known prior art patents and patent applications involving fractal antenna design include U.S. 20020190904 A1; U.S. Pat. No. 6,476,766; U.S. Pat. No. 6,452,553; U.S. Pat. No. 6,445,352; U.S. Pat. No. 6,140,975; U.S. Pat. No. 6,127,977; U.S. Pat. No. 6,104,349; WO 2004/001894; WO 03/023900; WO 01/54225; WO 01/54221; WO 99/57784; WO 97/06578; EP 1 313 166; EP 1 258 054; EP 1 227 545; EP 1 223 637 and ES 2 112 163. A list of known scientific publications is provided below at the end of the detailed description. Some of these publicly available documents also introduce the concept of space-filling curves. A space-filling curve is not a fractal, because it does not replicate itself in smaller scale. However, much like many fractals, space filling curves are defined by recursive replacement rules. There is a certain degree of similarity between the recursive iterations when a space-filling curve is developed. By proceeding through a large number of iterative replacement rounds it is mathematically possible to make a space-filling curve fill in a given space up to any given arbitrary percentage. A mathematically more accurate description of a genuine space-filling curve is a function that continuously maps the unit interval onto a bounded region of higher dimension.
The problems of known fractal and space-filling antennas are usually related to modest efficiency and too narrow bandwidth. Efficiency problems can be tracked to the requirement of making the meandering conductive trace in the antenna relatively long, in order to achieve an impedance match to the antenna port of a transceiver or receiver at required operating frequencies.
An objective of the present invention is to present an antenna that is small in size but still efficient enough for use in a portable communications devices. An additional objective of the invention is to ensure that such an antenna has a wide enough bandwidth. Another objective of the invention is to present an organized method for designing antennas of the kind meant above so that they match certain predefined criteria related to bandwidth, input impedance and efficiency.
The objectives of the invention are achieved by designing a radiating antenna element to resemble a space-filling curve of which certain non-contributing sections are eliminated.
According to an aspect of the invention there is provided an antenna for communication through radio frequency signals, comprising a radiating antenna element which is a meandering conductive line, wherein the meandering conductive line has the form of a pruned space-filling curve, in which a straight line segment exists at a location where a genuine space-filling curve would contain a bend.
According to another aspect of the invention there is provided a portable communications device for communication through radio frequency signals, comprising:
According to another aspect of the invention there is provided a method for manufacturing an antenna for communication through radio frequency signals, comprising the steps of:
According to another aspect of the invention there is provided a method for manufacturing an antenna for communication through radio frequency signals, comprising the steps of:
The invention is based on the insight according to which basic meandering and space-filling curves include certain sections that together produce an essentially zero net effect on the far field, if the curve is used as an antenna. Said zero net effect is a consequence of currents of essentially the same absolute magnitude flowing into essentially opposite directions in sections that are relatively close to each other. On the other hand, currents flowing through said sections give rise to reactive near fields, which in turn cause dielectric losses in the nearby dielectric materials. Also losses in the conductive material of the antenna itself may amount to not insignificant values, especially if the end-to-end length of the antenna is large. All in all, said sections can be considered as unnecessary, or even harmful from the viewpoint of the overall performance of the antenna.
The invention involves also a surprising observation according to which eliminating said unnecessary or harmful sections does not change the resonance frequency characteristics of the antenna even nearly as much as could be expected by simply looking at the decreasing end-to-end length of the antenna. Eliminating said unnecessary or harmful sections means deleting them from the basic or genuine meandering or space-filling curve and connecting the free ends of the remaining parts of the curve to each other in the most straightforward way. Since the new connection between said free ends is inevitably shorter than the original connection that included said unnecessary or harmful sections, the elimination makes the antenna shorter in end-to-end length. However, we have observed that as a result of eliminating the unnecessary or harmful sections, the resonance frequency of the antenna will only increase by a fraction of the percentage by which the end-to-end length decreased.
According to the invention, an antenna element is designed and manufactured to resemble a pruned meandering or space-filling curve. Conceptually the manufacturing process can be regarded to comprise generating a basic or genuine meandering or space-filling curve and performing an optimization calculation, in which sections of the basic or genuine meandering or space-filling curve are consecutively eliminated until a simulation calculation shows that a set of predefined operational criteria are met. A conductive antenna element is manufactured to match the meandering or space-filling curve after eliminating said sections.
The novel features which are considered as characteristic of the invention are set forth in particular in the appended claims. The invention itself, however, both as to its construction and its method of operation, together with additional objects and advantages thereof, will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings.
a to 2c illustrate the known generation of a space-filling curve through recursion,
a to 3e illustrate schematically some known space-filling antennas,
a and 4d illustrate the concept of eliminating segments or pruning,
a and 5b illustrate pruning a certain space-filling curve,
a to 6c illustrate pruning another space-filling curve,
a and 9b illustrate an antenna according to an embodiment of the invention,
a and 11b illustrate portable communications devices according to embodiments of the invention.
The exemplary embodiments of the invention presented in this patent application are not to be interpreted to pose limitations to the applicability of the appended claims. The verb “to comprise” is used in this patent application as an open limitation that does not exclude the existence of also unrecited features. The features recited in depending claims are mutually freely combinable unless otherwise explicitly stated.
In order to enable fully understanding the invention, certain known facts of monopole antennas and space-filling curves are first discussed.
The lowest operating frequency f1 of the straight wire monopole antenna corresponds to a wavelength λ1, for which h=λ1/4. When an oscillating signal of frequency f1 is applied to the antenna, the distribution of electric current along the length of the radiating antenna element 101, the length considered in the direction from the feed point 103 towards the open end of the radiating antenna element 101, is proportional to one quarter of a cosine wave with a maximum value at the feed point 103 and a zero at the open end. The two next highest operating frequencies f2 and f3 are odd integral multiples of f, (f2=3f1 and f3=5f1) and correspond to wavelengths λ2 and λ3, for which h=3λ2/4 and h=5λ3/4 respectively. At frequency f2 the current distribution along the length of the radiating antenna element 101 is proportional to three quarters of a cosine wave, and at frequency f3 to five quarters of a cosine wave respectively. At these operating frequencies a maximum of the current distribution is always located at the feed point 103, and a zero at the open end of the radiating antenna element 101.
a, 2b and 2c illustrate the principle of constructing a genuine or basic space-filling curve through iteration. The Hilbert curve is shown as an example.
In the following we will use the designation “space-filling antenna” to describe an antenna structure that otherwise resembles that shown in
a to 3e illustrate schematically various space-filling antennas. All of them have a ground plane 102 and a feed point 103. As a radiating antenna element, the antenna of
a to 3e should be understood schematically, so that they do not e.g. limit the mutual physical locations of the radiating antenna element and the ground plane. Later in this description we will consider physical locations of the various elements of the antenna structure, as well as potential couplings between the radiating antenna element and the ground plane, in more detail.
For the purpose of comparing with e.g. the linear monopole antenna we define the length l of the radiating antenna element in a space-filling antenna to be the physical length, measured along the curve, between the feed point and the point of the curve most distant from the feed point. It should be noted that for loop-shaped, ungrounded radiating antenna elements this means that the length is one half of the whole length of the curve. Due to the tightly meandering nature of the space-filling curve, said length is in all cases much greater than the height h of the radiating antenna element, or more generally the overall outer dimensions of the radiating antenna element.
For given ground plane dimensions and a given antenna height h it is easy to make a space-filling antenna have a much lower operating frequency than a straight wire monopole, simply because the length l of the radiating antenna element in the space-filling antenna is much longer than h. Conversely, for a given operating frequency, a space-filling antenna can be easily made to have a lower antenna height h than a straight wire monopole. Increasing the degree of recursion in the space-filling antenna further increases the length of the radiating antenna element (which in any case is typically much larger than λ/4) and correspondingly lowers the operating frequency. However, increasing the degree of recursion also tends to increase the level of losses.
A known characteristic of space-filling multiband antennas is that the operating frequencies are closer together in relative sense than those of a straight wire monopole. For example, a space-filling antenna having a radiating antenna element shaped like a Hilbert curve, the ratio of the second operating frequency to the first one is 2 or less depending on the degree of recursion, whereas for a straight wire monopole it is 3. During the development work leading to the present invention an exemplary set of space-filling antennas was measured. Said antennas all had identical ground planes and the same antenna height h. Each had a radiating antenna element shaped like a Hilbert curve, so that for the first antenna the degree of recursion was one, for the second antenna the degree of recursion was two, for the third antenna the degree of recursion was three and for the fourth antenna the degree of recursion was four. In a measurement between 200 MHz and 2 GHz the first antenna had one operating frequency band centered at approximately 1450 MHz and the second antenna had one operating frequency band centered at approximately 1200 MHz. The third antenna had two operating frequency bands at approximately 900 MHz and 1650 MHz, and the fourth antenna had a total of four operating frequency bands at 780 MHz, 1240 MHz, 1490 MHz and 1910 MHz.
a illustrates a piece of conductive line 401, which comprises segments A, B and C. Of these, segments A and C are parallel to each other and equal in length. If an electric current flows through the conductive line 401, with no component currents branching off or being added between said segments, the current IA flowing through segment A is opposite in direction but essentially equal in magnitude with the current IC flowing through segment C. Observed at a distance that is large compared to the distance between segments A and C and their length, the electromagnetic field effects caused by the currents IA and IC essentially cancel each other. Therefore, again observed at said relatively large distance, it would be difficult to tell the electromagnetic field effects caused by currents flowing in the conductive line 401 from those caused by currents flowing in the conductive wire 411 of
The difference between the conductive lines 401 and 411 is that the lafter is a “pruned” version of the former, where pruning is taken to mean eliminating segments that together cause a zero net effect when observed at a large distance—meaning segments A and C in
Assuming that the conductive lines 401 and 411 were made of the same material of identical thickness, located in identical surroundings, and also in all other ways similar to each other except for the elimination of segments A and C in the case of conductive line 411, it is easy to understand that an electric current of some identical value passing through each of them in turn will cause higher resistive and dielectric losses in the case of conductive line 401 than in the case of conductive line 411. The reason is the longer end-to-end length of the conductive line 401, which results in higher end-to-end resistance and larger electromagnetic interaction with the surrounding dielectric materials.
c and 4d illustrate a slightly more complicated case, in which a meandering section of the conductive line 421 of
The groupwise consideration of segments can be further generalized so that in pruning, a bend of arbitrary form in a meandering line of a genuine space-filling curve can be replaced with a straighter connection, if the result of a vector integral of the current distribution over said bend is closer than a predetermined limit to the result of a vector integral of the current distribution over said straighter connection.
For the purpose of evaluating the effects of pruning to the usability of the resulting curves as radiating antenna elements, we may briefly consider the mathematical modelling of an antenna, more exactly the method of moments (MoM) solutions to the boundary integral equations for antennas. In the method of moments, a radiating antenna element is considered to consist of a sequence of simple line segments. The current flowing through each segment is designated separately as an unknown variable, and these unknown variables are collected into a vector I. A system of linear equations is formed as
ZI=U (1)
where Z is the impedance matrix, and the voltage vector U contains the imposed input voltages. A common approximation regarding the voltage vector is that the incident voltage is localized to that segment of the radiating element that is closest to the feed point, which simplifies U so that it only contains one non-zero element. There are as many unknowns in the system of equations (1) as there are segments, or calculational elements, in the model of the radiating antenna element.
The diagonal elements of Z are called the self-impedances and they correspond to the impedances of the individual elements in free space. The non-diagonal elements of Z are called mutual impedances and they describe the interaction of the various calculational elements with each other. The exact values of the mutual impedances depend on the distances, sizes and relative orientations of the elements.
Let us suppose that the antenna is fed at the element number 1. We may compute the input impedance Zin of the antenna by setting the first element of the voltage vector U equal to some known input voltage U1 and all other elements of the voltage vector U equal to zero. Solving the system of linear equations gives the current distribution I of the antenna. We may write Zin=U1/I1 and, taken the formula for U1 from equation (1), expand as
where we have assumed that there are N segments in the model of the radiating antenna element. It is easy to interpret equation (2) so that in general the n:th term of the summation on the right-hand side gives the contribution of the n:th segment of the radiating antenna element to the overall input impedance, where n gets values from 1 to N.
If the conductive line 401 of
We may make the following assumptions and deductions:
1) Segments A and C are very close to each other in the sequential order of segments, which means that the currents IA and IC are of essentially the same absolute magnitude.
2) Segments A and C have the same length and direction, and are located far away from the feed point, which means that the mutual impedance terms Z1A and Z1C are of essentially the same magnitude.
3) As a consequence of assumptions 1) and 2) above, as well as of the fact that the currents IA and IC flow into exactly opposite directions, the terms related to segments A and C cancel each other from the summation.
4) Segment B is also far away from the feed point, which means that the mutual impedance term Z1B related thereto changes only little even if segment B is moved to the position shown as B′ in
As a general conclusion of the above analysis of
On the other hand, the principle of eliminating segments of a radiating antenna element can be generalized to cover more than two segments simultaneously. We may assume that a group of segments can be identified, for which the following assumptions hold to a reasonable accuracy:
1′) The sum of the moments, i.e. currents times lengths in vector representation, calculated over all segments in the group is zero, meaning that their net effect to the far field is zero.
2′) The sum of terms of the form Z1nIn/Il over all segments n of the group is zero, meaning that their net contribution to the input impedance is zero.
3′) Removing the segments of the group and correspondingly moving the remaining segments m causes only small changes to the mutual impedance terms Z1m corresponding to the remaining segments.
As a consequence the segments of the identified group can be removed without essentially changing the antenna's far-field behaviour or input impedance. In practical cases, the “reasonable accuracy” clause means that something “being zero” means that said something is close to zero than some predetermined, small limiting value.
a to 6c illustrate applying the pruning concept to two variations of the Hilbert curve, each time observing the conditions 1) to 4) or 1′) to 3′) above. The curve 501 of
b illustrates also schematically the possibility of making a grounding connection 513 at some carefully selected point along a radiating antenna element shaped like a pruned space-filling curve. Such grounding connections are used for tuning, and their coupling to the radiating antenna element and/or to the ground plane may be capacitive or galvanic. Also controllable switches may be used in grounding connection(s), so that selecting the state(s) of the switch(es) will dynamically affect the resonance characteristics of the antenna.
The Hilbert curve 601 of
Pruning, which can also be designated as removing segments that have been found to fulfil the conditions 1) to 4) or 1′) to 3′) above, has several benefits. Firstly, it makes the radiating antenna element simpler and thus easier to manufacture. It also makes the radiating antenna element shorter in length, which makes resistive losses slightly smaller. Additionally it makes dielectric losses smaller, because before pruning the small bends involved caused electromagnetic energy to be stored in the near fields of the bends, which made the antenna more susceptible to dielectric losses in the dielectric materials surrounding the radiating antenna element.
It has been found that even if pruning makes the radiating antenna element shorter in end-to-end length, it does not automatically increase the operating frequencies as much as could be expected. In an experiment made during the research work that led to the invention, pruning a radiating antenna element based on the Hilbert curve shortened the end-to-end length of the radiating antenna element by 35%, but only made the operating frequency 12% higher. In the process of designing an antenna this can be accounted for by first designing a space-filling antenna for which a simulation calculation shows the operating frequency to be somewhat too low, and then pruning until a renewed simulation calculation shows that the desired operating frequency has been reached.
When a low enough initial operating frequency has been obtained, there follows some pruning at step 704. The action taken at step 704 is described in more detail below in association with
At step 804 the input impedance of the antenna is recalculated with the elimination performed at step 803 taken into account. At step 805 a check is made, whether the change in input impedance that resulted from the elimination at step 803 is smaller than an acceptability threshold defined earlier at step 701. The check made at step 805 may take into account the one-time change in input impedance and/or an accumulated change since the pruning started. A positive finding at step 805 allows accepting the elimination according to step 806. If the finding at step 805 was negative, there follows a check at step 807, whether all possible pairs (or groups) of segments viable for elimination have been tried already. If not, there occurs a transition back to step 803 where another pair (or group) of segments is now selected. A positive finding at step 807 means that no solution can be found to the given design problem with the currently valid boundary conditions. In order to take into account the possibility of exiting step 704 through the failure-indicating substep 808 means that the process described in general in
The description has concentrated so far on single-band space-filling antennas. In case a dual- or multiband antenna is to be considered, the concept of finding an optimal antenna shape through pruning includes also the possibility of selecting, whether the pruning should affect only one operating frequency band or at least two operating frequency bands simultaneously. It should be noted that both impedance and current distribution depend heavily on frequency. If the relative magnitudes of at least two operating frequencies are to be kept the same, only such pairs or groups of segments should be selected for pruning for which the cancellation of currents and sameness of mutual impedance terms hold for all operating frequencies considered. On the other hand it is possible to change the multiband behaviour of an antenna by deliberately selecting such pairs or groups of segments for pruning for which the currents cancel each other at a first operating frequency but not at a second operating frequency. As a result, the input impedance after pruning stays the same at said first operating frequency but not at said second operating frequency, which effectively means a change in the second operating frequency. Equations (1) and (2) hold as such for each operating frequency in turn.
In the method diagrams of
a and 9b illustrate an exemplary antenna according to an embodiment of the invention. A basic support structure of the antenna is a dielectric plate 901. One surface of the dielectric plate 901 supports a radiating antenna element 902 in the form of a meandering curve, which has been obtained by pruning a space-filling curve. In this exemplary embodiment the curve 902 resembles closely that introduced previously in
For the sake of example,
The invention places few limitations for varying the structural solutions of the antenna. A non-exclusive list of possible variations is provided in the following. The support structure does not need to be planar or rigid; it can also be curved and/or flexible. Different kinds of support structures could allow at least a part of the ground plane to be placed on a plane that is perpendicular or at some other angle against some plane defined by the radiating antenna element. The radiating antenna element could extend onto two or more planes, or be genuinely three-dimensional. The unbalanced antenna structure could be replaced with a balanced one, making e.g. two space-filling curves constitute a di-pole antenna and using appropriate balanced feed systems. The line width of the radiating antenna element does not need to be constant. The ground plane could be partly or completely one upon the other with the radiating antenna element.
One possible generalization concerns the space-filling nature of the curves that are used as a starting point for designing antennas according to the invention. In the foregoing we have relied completely on space-filling curves. To be quite exact, the concept of optimizing an antenna through pruning as shown in
One possible area of applying the invention is the provision of an FM reception antenna to a portable communication device that also has important functionality on significantly higher frequencies. Portable communications devices that have evolved from what used to be just cellular telephones usually communicate with a cellular network on frequencies that are in the range from 800 MHz to 2 GHz. Antennas that work well with those frequencies are not applicable for reception on FM broadcasting frequencies, so a separate antenna should be provided for FM reception, if the same device is to additionally include an FM radio receiver. An antenna according to the invention is a good candidate for such an FM reception antenna, because the invention allows making it small and yet efficient, and because necessary structural factors such as dielectric support plates and ground planes typically already exist in a portable communication device.
a illustrates schematically a portable communications device 1101, which comprises a cellular communications part 1102 for communication with a cellular radio network. The antenna 1103, which is an antenna according to an embodiment of the invention, is connected to said cellular communications part 1102 for at least one of receiving radio signals from said cellular radio network and transmitting radio signals to said cellular radio network.
“Frequency independent features of self-similar fractal antennas”, Werner, D. H.; Werner, P. L.; Ferraro, A. J., Antennas and Propagation Society International Symposium, 1996. AP-S. Digest Volume: 3 1996, Page(s): 2050-2053.
“Frequency independent features of self-similar fractal antennas”, Werner, D. H.; Werner, P. L., Radio Science Volume 31, No. 6, 1996 Pages 1331-1343.
“Fractal constructions of linear and planar arrays”, Werner, D. H.; Haupt, R. L., IEEE Antennas and Propagation Society International Symposium, 1997, 1997 Digest Volume: 3, 1997, Page(s): 1968-1971
“Fractal antenna engineering: the theory and design of fractal antenna arrays”, Werner, D. H.; Haupt, R. L.; Werner, P. L., IEEE Antennas and Propagation Magazine Volume: 41, No. 5, October 1999, Page(s): 37-58.
“Radiation characteristics of thin-wire ternary fractal trees”, Werner, D. H.; Rubio Bretones, A.; Long, B. R., Electronics Letters Volume: 35, No. 8, 15 Apr. 1999, Page(s): 609-610.
“The theory and design of fractal antenna arrays”, Werner, D. H.; Werner, P. L.; Jaggard, D. L.; Jaggard, A. D.; Puente, C.; Haupt, R. L.; In: “Frontiers in Electromagnetics”, Werner, D. H., Mittra, R., editors., IEEE Press, New York, 2000. Pages 48-93.
