I have already reported twice on the Core Wireless v. LG litigation in France: the first time to report on the first instance judgment, and the second time to report on a procedural ruling issued during the appeal proceedings.
Today’s post will be about Conversant v. LG… which is exactly the same litigation, since Core Wireless Licensing SARL has now become Conversant Wireless Licensing SARL.
So, the appeal ruling is out – and it basically confirms the first instance judgment.
As a reminder, Conversant (Core) holds a portfolio of Nokia-originating telecom patents. More than a thousand of these patents were declared essential for the 2G, 3G and 4G standards before the European Telecommunications Standards Institute (ETSI).
Core negotiated for several years with South-Korean giant LG Electronics in an attempt to reach a license agreement over this portfolio – but the negotiations failed. In September 2014, Core filed a complaint with the Paris Tribunal de grande instance (TGI), claimed damages and requested that the court should set a royalty rate for LG’s telecommunication devices. The complaint was based on a set of five European patents.
The TGI held that there was no evidence of infringement of the five patents by LG, and thus dismissed Core’s claims in April 2015.
Core appealed. When I commented on an interim case management ruling, I assumed that the only remaining issue on appeal was the determination of the royalty rate. This understanding was based on a somewhat misleading comment in the interim ruling – and it was thus incorrect.
In fact, similar issues were raised on appeal as in first instance. The main difference is that Conversant no longer asserted three of the patents initially discussed and preferred to focus on the remaining two.
Conversant still claimed that these two patents were infringed by LG and requested that the court should set a worldwide FRAND royalty rate for its portfolio. LG raised a number of defenses: invalidity of the two patents, absence of FRAND offer by Conversant, exhaustion of rights (based on a Nokia – Qualcomm license), absence of essentiality of the two patents, etc.
The first interesting part of the appeal ruling is a procedural one. The parties and the court made extensive use of the new provisions offered by the law of July 30, 2018 on the protection of trade secrets.
As already explained in the report on the interim ruling, the access to some of the exhibits (especially license agreements) was restricted to the parties’ lawyers and to some designated persons having signed confidentiality agreements (especially interpreters and economist experts).
Also, two versions of the written submissions, were filed, namely a complete one and a redacted one deprived of any reference to confidential information relating to the various license agreements at stake.
Finally, the hearing was scheduled over three days.
On the first day, Nokia’s FRAND commitment, an alleged abuse of dominant position and the exhaustion defense were debated. During part of the day, access to the courtroom was restricted to the lawyers of the parties as well as a couple of representatives from said parties. This closed session revolved around the most sensitive exhibit (the Nokia-Qualcomm agreement) and the exhaustion defense.
On the second day, the determination of the FRAND rate was discussed. Again, access to the courtroom was restricted to the above persons, plus a number of designated experts, during part of the day – when comparable license agreements were disclosed.
On the third day, validity, essentiality and infringement of the patents were discussed (without any access restriction).
The recent modernization of French procedural rules which has made this type of customized arrangement possible is most commendable – and was long overdue. It has probably put France on an equal footing with other major patent jurisdictions that have long been accustomed to such notions.
As readers may already have noted, a three-day appeal hearing for a patent case, especially at the appeal stage, is truly exceptional in France. This duration was certainly justified by the complexity of the case.
Hence, it could legitimately be assumed that a significant decision would come out of this whole process, namely a worldwide FRAND rate-setting determination. It did not.
First, the court reviewed the two sample patents asserted by Conversant, and addressed the invalidity challenges brought by LG. The court concluded that both patents were valid.
Second, the court addressed whether these two patents, declared as essential, were indeed essential for the standards at stake. Walking in the footsteps of the first instance judges, they concluded that they are not.
Since the demonstration of infringement was solely based on the essentiality of these patents (and on the compliance of LG’s devices with the standards), and since no other patent from the portfolio was asserted in the lawsuit, the court was finally not aware of any actual Standard-Essential Patent (SEP) held by Conversant.
As a result, the court did not rule on exhaustion of rights, on the FRAND royalty rate and on all other exciting questions raised by the case.
Again, I guess that a number of stakeholders may be disappointed by this underwhelming decision, not so much because the plaintiff lost, but because legal certainty is currently relatively low when it comes to SEPs, FRAND royalty rates and exhaustion of rights.
Additional guidance from a French court would thus have been welcome.
That being said, looking at the non-essentiality finding in the ruling in more detail is rather interesting.
Claim 1 of the first patent, EP 0978210, is directed to a “method for selecting a base station in a mobile communication system“. One of the steps of the method consists in, “in response to recognizing the poor radio signal, measuring“, well, a certain something.
According to the relevant technical specifications of the UMTS (3G) and LTE (4G) standards, as explained by the parties and their experts, if a radio connection having a quality lower than a threshold is identified, a measuring step as claimed must be performed. But if the quality of the radio connection is higher than the threshold, the terminal may decide not to perform any measurement, which means that it may conversely decide to perform a measurement anyway.
The court concluded that a terminal which also performs the measurement when there is a good connection would be compliant with the standards but would not infringe the patent. Therefore, the patent is not an SEP.
Claim 1 of the second patent, EP 0950330, can be quoted in its entirety: “A user terminal comprising a wireless interface and a formatting device for formatting a signal for transmission over said wireless interface in accordance with a low level signal format protocol, characterised in that it comprises means for receiving a type signal, wherein said type signal is indicative of a high level signalling protocol to be used in transmitting said signal, and for formatting said signal in accordance with said high level signalling protocol“.
According to Conversant’s expert, LTE terminals can access both of the so-called IPv4 and IPv6 networks. If a network can work with only one of IPv4 or IPv6, the terminal will use it. If a network can work with both IPv4 and IPv6, the terminal will select one of those. Thus, the claimant argued, the subject-matter of claim 1 is implemented.
However, LG replied that IPv4 and IPv6 are just two versions of the same protocol, and not two different protocols. The court was quite sensitive to this argument, as the patent itself, although it was filed after IPv4 and IPv6 were established, does not distinguish between these versions and only broadly refers to the internet protocol (IP). Besides, the standard does not explicitly state that the terminal should select between IPv4 and IPv6 when both are available. It is thus not possible to consider that such a selection step would be demanded by the standard.
Again, the conclusion is that the patent is not an SEP.
Essentiality was thus the core issue throughout this litigation.
This appeal ruling should come as a warning that claim mapping of a standard is far from being straightforward.
Among the 1,000+ patents of the portfolio, it can be safely assumed that the five (later narrowed down to two) patents asserted in this case were rated as being among the “strongest” or “best” ones by their holder. And yet, even with these two particularly “strong” or “good” patents, the essentiality case was not clear-cut, as delicate issues of claim interpretation arose.
This is not completely surprising, as standards are elaborated quite independently from the patents which may be owned by members of the standard-setting body. Also, a standard contains a number of minimal rules but does not provide all implementation details. Even if, in practice, all implementations use a certain feature, it does not necessarily mean that this feature is required by the standard itself.
This makes me wonder about the proportion of patents declared as potentially essential which would be viewed as actually essential by a court? I assume that proportion could in fact be rather low. Definitely some food for thought for all those involved in license negotiations.
CASE REFERENCE: Cour d’appel de Paris, pôle 5 chambre 1, April 16, 2019, Conversant Wireless Licensing SARL v. LG Electronics France SAS & LG Electronics Inc., RG No. 15/17037.