Wednesday 25 July 2012

T 667/08 – Looking Everywhere


This is an appeal against the refusal of a divisional application.

Claim 1 of the sole request before the Board read:
A monopole antenna, said monopole antenna comprising a radiating arm and a ground plane (12), said radiating arm being excited by a transmission line (11),
wherein said radiating arm is shaped as a space-filling curve (59-60),
wherein said space-filling curve is composed by at least ten connected segments forming a non-periodic portion of said curve, wherein: 
  • each of said segments is shorter than a tenth of the operating free-space wave length of the antenna; 
  • said segments are spatially arranged in such a way that none of said segments form, together with an adjacent segment, a longer straight segment;  
  • said segments are connected in such a way that each segment forms an angle with its neighbours; 
  • none of said segments intersect with another of said segments except at the ends of the curve, whereby said space-filling curve intersects itself at its beginning and end so that said space-filling curve forms a closed loop; 
  • each pair of adjacent segments of said curve forms a corner; and 
  • wherein, if said curve is periodic along a fixed straight direction of space, the corresponding period is defined by the non-periodic portion composed by at least ten connected segments, none of said connected segments forming, together with an adjacent segment, a straight longer segment;
said space-filling curve being a curve that features a box-counting dimension larger than one;
wherein said curve is not self-similar.
The Board found this request to be allowable.

In the following extracts, the Board deals with the question of whether A 123(2) and R 27(1)(e) EPC 1973 (now R 42(1)(e)) had been complied with.

A123(2)

[4.1] In this section, references to the original disclosure or original claims refer to the original version of the present divisional application as published under No. 1 592 083. It is stressed that this version is identical to the original earlier application.

[4.1.1] The present application relates to “Space-filling miniature antennas”. In its original version, the description underlines that a key point of the invention resides in the definition of a “novel geometry, the geometry of Space-Filling Curves (SFC)” which is used to shape a part of an antenna […]. The original disclosure consistently stresses the general benefit in terms of size reduction which is expected from the use of SFCs when designing antennas or, given a fixed size, in terms of the ability of the antenna to operate at lower frequencies with respect to a conventional antenna of the same size […]. The original disclosure neither provides much detail as to antennas according to the various embodiments described throughout the description nor does it put any particular emphasis on the advantages they offer. In this respect, the dipole antenna, the monopole antenna, the patch antenna, the slot antenna, the loop antenna, the horn antenna and the reflector antenna, to which explicit reference is made in the original application, constitute mere illustrations of possible implementations of the original idea of using SFCs for the design of common antenna types.

The original application discloses various examples of SFCs. In this respect, no clear distinction is made between the various categories of SFCs, in particular, between self-similar curves (i.e. structures which provide repetition in smaller and smaller scales) and non self-similar configurations. As a matter of fact, the original disclosure does not even expressly refer to these different categories of SFCs. Some of the geometries originally considered are “inspired in the geometries studied already in the XIX century by several mathematicians such as Giusepe Peano and David Hilbert. In all said cases the curves were studied from the mathematical point of view but were never used for any practical engineering application” [… M]ention is made of techniques which might have been used in order to construct SFCs. These include the use of algorithms known as Iterated Function System (IFS), Multireduction Copy Machine (MRCM) or Networked Multireduction Copy Machine (NMRCM). A combination of said various techniques is also envisaged […].

However, contrary to the appellant/applicant’s assertion, the European search report revealed the existence of various documents from which it followed that the use of SFCs for designing antennas was already known at the time of filing of the application from which a priority right is claimed. While a fractal pattern contributes to the inductive loading of the antenna system according to document WO-A-99/27608 (D2), the further documents WO-A-97/06578 (D1) and ES-A-2 112 163 (D5) describe antennas whose radiating arm is designed as an SFC. In both documents D1 and D5, the radiating portions of the antenna define self-similar (deterministic) curves.

Apart from original claim 6, which reads: “A monopole antenna comprising a radiating arm and a ground counterpoise in which at least a part of said is shaped either as an SFC, Hilbert, Peano, HilbertZZ, SZ, Peanoinc, Peanodec, PeanoZZ, or ZZ curve according to claim 1, 2, 3 or 4”, the only example of a monopole antenna in the original application documents is given in Figure 4 and the corresponding section of the description.

[4.1.2] In the Board’s view, the original disclosure provides a sufficient basis for the deletion in present claim 1 of the adjective “straight”, initially associated with the feature of the “at least ten connected segments” in the definition of the space-filling curve in original claim 1. It is observed, in this respect, that the passage of the original description in paragraph [0010], which provides in similar terms a definition of an SFC, does not include said adjective when referring to the “at least ten connected segments”. The additional limitation, in the same paragraph, according to which “no pair of adjacent segments defines a larger straight segment” does not necessarily imply that the segments, as such, are straight but appears to specify, in the case that these segments are indeed straight, that they cannot be associated so as to define a longer straight segment (cf. point 3.1 above). A further confirmation of the fact that the reference to segments was never intended to be limited to straight configurations can be found in the statements in paragraph [0010], according to which the SFC can be fitted over a curved surface, or in paragraph [0027], according to which a reflector whose perimeter is shaped with an SFC can be either flat or curved. As stressed by the appellant, at least some segments of a pattern that is fitted over a curved surface must themselves be curved.

[4.1.3] The original disclosure also provides sufficient basis for the claimed configuration of a monopole antenna whose radiating arm defines a closed loop. More specifically, original claim 6 defines a monopole antenna comprising a radiating arm and a ground counterpoise, the arm being shaped as an SFC. Claim 6 explicitly refers to original claim 1 in which the possibility of the segments of the SFC intersecting at the tips of the curve is explicitly acknowledged.

[4.1.4] A main issue to be considered under the aspect of added subject-matter concerns the definition of the SFC in claim 1, in particular the indication that “said curve is not self-similar”. More specifically, the question arises whether a sufficient basis exists in the original disclosure allowing a generalisation of the teaching of the monopole antenna of Figure 4. Whilst it is acknowledged that the SFC disclosed in this example is indeed not self-similar, this aspect is not emphasised in the section of the description relating to Figure 4 nor indeed in the rest of the application, where the terms “self-similar”, “self-affine” or “deterministic” are not even employed.


It is an undisputed principle in the jurisprudence of the Boards of Appeal of the EPO that an amendment is allowable under A 123(2) if the subject-matter resulting from the amendment is directly and unambiguously derivable from the original application documents i.e. the description, the claims and the drawings, using common general knowledge. Thereby it is not necessary that the subject-matter resulting from the amendment was explicitly disclosed in the original application.

It is therefore essential, when deciding on issues of added subject-matter, to identify the actual teaching conveyed by the original disclosure, i.e. the technical information that the skilled person reading the original disclosure would have derived from its content (description, claims, drawings) considered in its entirety.

This approach might lead to the identification of subject-matter which has not been explicitly revealed as such in the application as filed, but nevertheless derives directly and unambiguously from its content. Literal support is not required by the wording of A 123(2). An amendment can therefore be allowable if it combines information which has not been disclosed in one and the same section of the original disclosure, but results, for instance, from information gathered from various embodiments possibly associated with general statements regarding the information derivable from the introductory section of the application.

If this were not the case, the original disclosure would be deprived of a part of the information it actually contains, namely the technical teaching that the skilled person would retrieve from the application but which may typically extend beyond a mere literal interpretation of the original text.

In order to avoid any misunderstanding, it is stressed that the Board does not question the principle according to which embodiments of an invention can normally not be freely combined under A 123(2) but merely emphasises that each embodiment in a disclosure must be construed with the knowledge and understanding aptitude of the skilled person in the art in the light of the whole application.

Under the present circumstances, the broadest message conveyed by the original application is to design a part of an antenna as an SFC so as to limit its size or the area it occupies. The section “Background and summary of the invention” establishes that this idea can be implemented in various types of antennas and, in particular, in monopoles. In the absence of any restriction as to the type of SFC to be associated with each type of antenna, the skilled person would have understood that all types of SFC, as defined in paragraph [0010], could be implemented in any of the antenna types, since the effect in terms of size reduction resulted solely from shaping the antenna as an SFC irrespective of the type of antenna considered. This also applies inter alia to monopole antennas. This interpretation of the original description is further confirmed by the wording of original claim 6 which refers to a monopole antenna comprising a radiating arm shaped, for example, as an SFC. As convincingly put forward by the appellant, the skilled person is aware of the fact that the techniques of IFS and MRCM, referred to in paragraph [0009], generate self-similar structures. In contrast, the use of an NMRCM algorithm permits to avoid said self-similarity. In this context, reference was made to common general knowledge, as for example exemplified in the book: “Chaos and Fractals – New Frontiers of Science”, Heinz-Otto Peitgen et al., Springer-Verlag, 1992 (cf. in particular section 5.9 “Breaking Self-Similarity and Self-Affinity or Networking with MRCMs”).

For these reasons, the Board is convinced that the skilled person would have indeed recognised that the evocation of these various techniques actually constitutes an implicit reference to the two families of SFCs, namely those which are self-similar (or self-affine) and those which are not, bearing in mind that IFS is essentially equivalent to MRCM, at least insofar as the result is concerned (cf. Peitgen, Section 5.1).

Furthermore, the Board holds that the lack of a reference in the independent claims to the NMRCM algorithm, possibly associated with IFS or MRCM algorithms, to generate non self-similar curves as disclosed in original claim 16, is allowable. As pointed out by the appellant during the oral proceedings, the SFC disclosed in relation to Figure 4, or its reproduction in cascade in Figure 2 is not obtainable by NMRCM. The skilled person would have thus recognised that the monopole antenna actually disclosed in the original application was limited neither to the specific shape of Figure 4 nor by the technique of NMRCM to generate non self-similar structures but did in fact encompass possible alternatives. The skilled person would therefore have understood that what really mattered was just to break the similarity of self-similar patterns, independently of the technique required to reach this objective.

[4.1.5] In conclusion, the subject-matter of present claim 1 derives directly and unambiguously from the original disclosure. It therefore meets the requirements of A 123(2). […]

R 27(1)(e) EPC 1973 – Example embodying the invention

[5] A consequence of the amendments made in the course of the examination and appeal proceedings is that the claimed invention is devoid of any concrete example. It is pointed out, in this respect, that the sole illustration of a monopole antenna in Figure 4 depicts an open radiating arm contrary to the claim’s wording. The present Board (in a different composition) has, however, already decided in decision T 990/07 that it is not always necessary under R 27(1)(e) EPC 1973 to include an example embodying the invention if the earlier application is considered to fulfil the requirements of A 83 EPC 1973 (cf. T 990/07 [3]). In the present circumstances, the Board has no doubts that the realisation of a radiating arm defining an SFC which is not self-similar for the realisation of a monopole antenna can be carried out on the basis of present disclosure. It is observed, in this respect, that the printed circuit fabrication techniques referred to in relation to Figure 3 could be applied irrespective of the actual nature of the SFC considered. The requirements of R 27(1)(e) EPC 1973 are therefore met.

To download the whole decision, click here.

The file wrapper can be found here.

1 comments:

Anonymous said...

K's Law may be retired but it is still an excellent resource.

Thought I'd do my bit to add to it my pointing out that this decision made it into the September 2013 edition of the Guidelines for Examination, specifially Part H-IV, 2.3.