$1.5B Verdict Against Microsoft Overturned, New Trial Granted

On January 29, 2007, a jury trial commenced on issues pertaining to audio coding patents U.S. Patent Nos. 5,341,457 and RE 39,080 (“the ‘457 patent” and “the ‘080 patent,” respectively). On February 22, 2007, the jury returned a verdict in favor of Lucent finding the patents valid and infringed by Microsoft. Damages were assessed at $1.54B dollars.

Yesterday, in a 43-page opinion, Judge Rudi Brewster ruled on over 20 post-trial motions filed by Microsoft seeking to overturn the verdict, or to grant a new trial. On the issues of infringement and damages for the '457 patent, all motions were granted, With regard to the '080 patent, Microsoft's JMOL for infringement was denied, but Judge Brewster held that, since Fraunhofer is a co-owner of the '080 patent, Lucent lacked standing to bring the action, and Microsoft's license defense insulated them from liability.

Ownership and License of the '080 Patent:

Among the affirmative defenses raised by Microsoft, two related to the ownership of the ‘080 patent. Microsoft asserted that the ‘080 patent was co-owned by Fraunhofer Gesellschaft (“Fraunhofer”), a German research company. As such, Microsoft contended that Lucent lacked standing to bring suit against Microsoft for infringement of the ‘080patent. Microsoft also asserted the defense of license, contending that it had a license from Fraunhofer to practice the ‘080 patent.

Under a 1989 research agreement between AT&T and Fraunhofer (“the JDA”), the court found that Fraunhofer was a co-owner of the ‘080 reissue patent, since the JDA was never terminated:

The JDA between Fraunhofer and AT&T covered collaborative work during the period of the stay of Karlheinz Brandenburg at AT&T. This period began in April 1989 and ended in June 1990. All work done on digital audio coding during this period was classified as “New Work” and would be jointly owned by AT&T and Fraunhofer. In 1991, the period covering New Work was extended by AT&T and Fraunhofer to cover work that continued after the expiration of the JDA and the departure of Brandenburg. The parties agreed to extend the period indefinitely, until one of the parties gave notice to terminate the arrangement. No evidence was presented that the parties ever terminated the agreement. Hence, the combined effect of the JDA and the extension letter defines the period of New Work as beginning in April 1989 at the arrival of Brandenburg and continuing indefinitely thereafter.

Instead of partitioning the claims into separate applications, AT&T/Lucent decided lump together "New Work" claims (which traced back to Fraunhofer) with separately-developed claims. The court found this fatal to Lucent's case:

[B]y filing claims to New Work in the same application as claims to Existing Technology, the patent is jointly owned by AT&T and Fraunhofer. AT&T cannot cause Fraunhofer to forfeit its ownership rights to New Work simply by filing AT&T’s Existing Technology in the same patent application. If anything is forfeited, it is AT&T’s rights to exclusive ownership of the entire ‘938 patent. AT&T forfeited that right when it chose to file all four claims in one application. Lucent made a similar decision when it applied for the reissue that became the ‘080 patent and chose to keep claim four in the same application with claims one and three, all with one priority claim. The Court therefore FINDS that under the JDA, the ‘938 patent and its reissue, the ‘080 patent, are jointly owned by AT&T and Fraunhofer.

Since Microsoft had already obtained a license from Fraunhofer, "Microsoft as a licensee cannot be liable for infringement of the ‘080 patent. 35 U.S.C. § 271."

Infringement of the '457 Patent:

According to the court, there was no direct evidence that Microsoft's HQ encoder performed the patented methods presented at trial. Even Lucent’s expert testified that he had never observed the HQ encoder running, he had not made a CD encoded by the HQ encoder, and he could not explain how a user would select and run the HQ encoder. Instead, Lucent argues that its
proof of infringement centered on circumstantial evidence that the HQ encoder would run automatically when the fast encoder failed and therefore carry out the claims of the ‘457 patent.

After reviewing the record, the court ruled that the circumstantial evidence was not enough to show infringement, since too much speculation existed over the actual operation of the HQ encoder:

In essence, Lucent offered circumstantial evidence that proved at most that the HQ encoder was possibly capable of running. What it failed to show, circumstantially or otherwise, was that HQ had actually ever run and performed the claimed method. This evidence is insufficient as a matter of law to demonstrate infringement of a methods claim. While Lucent argues that “the record was replete with evidence that the HQ encoder will be invoked as a result of common conditions that occur routinely in practice,” as the Federal Circuit noted in E-Pass Technologies, Inc. v. 3Com Corp., 473 F.3d 1213, 1223 (Fed. Cir. 2007), if it was so common and so routine, certainly Lucent could have introduced evidence of at least one instance where the HQ encoder had run.

KSR Changed Everything? Not Really:

Microsoft tried dangling KSR before the court, but Judge Brewster wouldn't bite:

[T]he Court is unpersuaded that KSR’s “broad implications” warrant reopening discovery and searching for new prior art. Although Microsoft points out that this Court allowed reopening of discovery in Groups 4 and 5, the circumstances differ here. Groups 4 and 5 had not started trial before the KSR ruling came out, whereas the Group 2 trial has been completed. Additionally, an examination of district court and Federal Circuit rulings on obviousness since the KSR decision does not support Microsoft’s requested upheaval. In sum, Microsoft has not offered any meritorious reason why obviousness on the ‘080 patent should be re-tried or why the jury’s
verdict as to non-obviousness of the ‘080 patent was not support by sufficient evidence. Therefore, the motions for judgment as a matter of law and for a new trial on obviousness are DENIED.

Damages and the Entire Market Value:

The court found two major problems in applying the entire market value rule to Microsoft. The first was the failure of the evidence to establish a link between the cost of the computers (rather than the operating system, Windows Media Player, the MP3 codec or some other “unit”) and the customer demand or value of the patented technology. The second "and probably even more
troublesome problem is the failure to establish that the patented features themselves produced any customer demand or value of the product."

Here, the Court finds no evidence adduced at trial that establishes that the patented features of either the ‘457 or the ‘080 patents were critical to MP3 or that they established the basis for the customer demand or value of MP3, let alone were critical or provided value to the whole computer. The evidence cited by Lucent from the trial record shows only that MP3 capabilities overall were a commercially important feature. According to Lucent’s expert Dr. Jayant, MP3 technology originated from many sources and the MP3 standard specifies many aspects of audio compression. Although Jayant pointed out what he considered important in the MP3, the key technology he identified related to the decoding of the bit stream syntax (which allows decoders from different manufactures to interpret and play back the audio signal); he did not identify either the invention of the ‘080 patent or the ‘457 patent as critical to MP3. Additionally, the evidence demonstrated that although the inventions of these patents could be used with the MP3 standard, they were not required or critical to practice the MP3 standard.

Royalty Rate:

With regard to the royalty rate, the court denied Microsoft's JMOL, but granted a new trial, noting that

[I]t is not so much that the jury lacked any evidence on which to base a reasonable royalty rate, but that the evidence provided limited guidance . . . although a plethora of licensing agreements were admitted into evidence, the majority of these which advocated a royalty rate in the 0.5% range lacked sufficient relevance to the technology at issue here, the relevant date of the hypothetical negotiation and/or the scope of a license that would be negotiated between these parties.

View/download the opinion here (link).