Tuesday, September 25, 2018

Qualcomm says FTC motion aims to "radically reshape [standard-essential patent] licensing in the cellular industry"

The two most important motions for summary judgment that Qualcomm is facing at present (and one might even say has ever faced in its history) are the Federal Trade Commission's motion to hold that Qualcomm itself committed to extending standard-essential patent licenses to rival chipset makers such as Intel (a motion that has drawn broadbased support from industry) and a motion by Apple and four contract manufacturers to end Qualcomm's double-dipping practice (selling chipsets and additionally collecting patent license fees). This post is about Qualcomm's opposition in the Northern District of California to the former, but I'd also like to mention that Qualcomm is trying to duck the latter by means of a motion to dismiss all declaratory judgment claims relating to Qualcomm patents from the Apple (and contract manufacturers) v. Qualcomm case in the Southern District of California. When it turned out that its adversaries were going to insist on an adjudication of their patent exhaustion defense, Qualcomm requested expedited briefing, which Judge Gonzalo Curiel denied. The opposition brief to Qualcomm's attempt to chicken out is due by the end of next week.

Back to San Jose. Qualcomm's opposition to the FTC motion is underwhelming. It's unlikely that Judge Lucy Koh will agree with Qualcomm's positions on the core issues; at the most--and even that part is doubtful--the judge might find that some evidence needs to be discussed at trial.

Here's the opposition brief, and below the document I'll explain why I still believe the FTC motion is perfectly meritorious:

18-09-24 Qualcomm Oppositio... by on Scribd

In the first paragraph, Qualcomm acknowledges the importance of the issue the FTC motion is about:

"Under the guise of seeking merely to 'streamline the trial' [...], the FTC's Motion for Partial Summary Judgment requests a ruling that would radically reshape licensing in the cellular industry, not just for Qualcomm but for all cellular SEP holders that made FRAND commitments to ATIS or TIA. The FTC contends that those SEP holders have a clear and unambiguous contractual obligation to grant licenses to manufacture and sell modem chips."

Still on the first page, Qualcomm's brief accused the FTC of attempting "a complete reimagining of the industry."

Qualcomm points to the field-of-use limitation found in standard-essential patent (SEP) licensing commitments. The idea is that you can request a FRAND license only for the purpose of implementing the relevant standard. The reason is simply that one could otherwise request a license in order to use patents in a competing standard, or for non-compliant products. But Qualcomm claims that only "complete devices" can implement cellular standards. Qualcomm more or less concedes (or at least does not deny with specificity) that some SEP claims may be practiced by baseband chips. However, it argues that other components of a mobile device are also required. Examples that Qualcomm mentions in its discussion of standard-setting documents include "antennae" and "power-control chips."

There's no question that you need an antenna and electricity for mobile telephony. But that doesn't mean the mobile baseband chip--or "modem chip" in accordance with Qualcomm's brief--doesn't implement the standard simply because it's the central and decisive component. It's the mastermind (a term that is key to the analysis of an alleged "divided infringement" of a method claim and fits here, too). Once the importance of the mastermind component is downplayed, the ultimate consequence may be that not even a device infringes since electricity must be provided by a utility, which in turn needs some energy source somewhere.

In a 2014 case, GPNE Corp. v. Apple, Judge Koh herself held "as a matter of law that in [that] case, the baseband processor [was] the proper smallest salable patent-practicing unit." And in Judge Koh's court, a 2012 jury sided with Apple against Samsung on patent exhaustion, based on Samsung's license to Qualcomm and Apple's use of Qualcomm chips in certain products at issue back then. The same happened to Samsung in the Netherlands and France (where I attended a preliminary injunction hearing in 2011).

It's actually a cornerstone of Qualcomm's business model to position itself as a SEP clearinghouse. That's why Qualcomm's brief now seeks to resolve the inconsistency between what courts have found as a result of Qualcomm's licensing practices and what Qualcomm is claiming--and must claim--at this stage in its opposition to the FTC motion:

"Although Qualcomm has obtained inbound cross-licenses to the cellular SEPs of its licensees, those were to ensure that its licensees cannot sign a license for rights under Qualcomm's patents and then opportunistically assert their own patents against Qualcomm."

If Qualcomm's chips didn't implement those stnadards, how could other SEP holders "assert their own patents against Qualcomm" by suing device makers? Qualcomm then points to its practice of neither licensing nor suing rival chipset makers, but that doesn't mean anything. Just like Ericsson--the first company Qualcomm is pointing to in its opposition brief--Qualcomm simply does this for leverage. Almost five years ago I wrote about Ericsson's fairly honest concession of the whole rationale behind going after device makers, not chipset makers.

With regard to the ATIS and TIA FRAND promises at issue, Qualcomm claims that their non-discrimination requirements merely "prohibit an SEP holder from charging one licensee a 1% royalty while (all other terms being equal) charging a similarly situated licensee a 5% royalty." However, not extending a license at all--no matter the royalty rate--is the most extreme form of discrimination.

Qualcomm argues the ATIS and TIA FRAND declarations don't say what they say because those American standard-setting bodies must have patent licensing policies consistent with those of ANSI. The reason why this doesn't convince me at all is simple: Qualcomm signed those declarations and is bound to them regardless of whatever principles and obligations may govern the inter-association relationship between ATIS and TIA on one side and ANSI on the other side. If Qualcomm feels ATIS and/or TIA are out of compliance with ANSI's rules, it can ask ANSI to investigate and take whatever action as a result, but it doesn't alter the language of Qualcomm's declarations. Also, if the horse is out of the barn, that's it: even if ANSI didn't and doesn't require ATIS and TIA to ensure that all comers, including rival chipset makers, receive a FRAND license to certain SEPs, it doesn't mean ATIS and TIA couldn't do so, and all that the likes of Intel need is one valid and enforceable contractual basis. ANSI's policies don't say that it's against its policies to license chipset makers.

Qualcomm furthermore argues that the FTC's concession that the ETSI FRAND declaration can't be interpreted without considering evidence suggests the same applies to the ATIS and TIA declarations. But what makes the key difference here is that the ETSI declaration is under foreign (here, French) law. A Qualcomm expert report on ETSI is attach to the opposition brief as Exhibit 35. The PDF shown further above contains all public exhibits, such as this one.

Getting back to the question of whether baseband chips implemenet cellular standards: they do because they control the transport-layer communication between mobile devices and base stations. Qualcomm has its reasons (as do Ericsson and others) why it prefers not to license or sue chipset makers, but if it wanted to sue them, it could do so, and I believe it wouldn't even need to resort to indirect-infringement theories.

Share with other professionals via LinkedIn: