Cell Phone Network Patent Survives § 112 Attacks in Federal Circuit

By Chris Reaves

HTC Corporation v. IPCom GmbH & Co., KG, No. 2011-1004, Fed. Cir. Jan. 30, 2012

In a recent infringement suit, a patent for a cell phone network mobile station narrowly survived several attempts at invalidation through indefiniteness.  The Federal Circuit applied numerous rules of claim interpretation, especially for means-plus-function claims, and reminded the defense of the importance of addressing all issues at the trial level.  The case serves as a good lesson in “what not to do” for both prosecutors and litigants.

Background

IPCom GmbH & Co., KG owns US Patent No. 6,879,830 (‘830) and other patents, which go to cell phone networks and their methods.  ‘830 in particular discusses an efficient cellular phone network “handover” – the process by which one cell tower takes over a call from another as the phone moves from place to place.

IPCom and the HTC Corporation countersued each other for, respectively, infringement of three of IPCom’s patents, including ‘830, and declaration of non-infringement.  During the trial’s Markman hearing, HTC moved for summary judgment of invalidity of ‘830 claims 1 and 18.  The district court found that the means-plus-function portion of these claims had the required corresponding structure.  However, it also found that the claims included both method steps and an apparatus, making them indefinite under the rule of IPXL Holdings v. Amazon.com.  It therefore invalidated both claims.

On stipulation of the parties, the decision was promptly declared final and certified for appeal under Fed. R. Civ. P. 54(b).  Judges Bryson, Linn, and O’Malley of the Federal Circuit heard the appeal.

IPXL Indefiniteness Does Not Apply to “Preamble-Within-A-Preamble”

Claim 1 of ‘830 recites, in total:

A mobile station for use with a network including a first base station and a second base station that achieves a handover from the first base station to the second base station by:

storing link data for a link in a first base station,

holding in reserve for the link resources of the first base station, and

when the link is to be handed over to the second base station:

initially maintaining a storage of the link data in the first base station,

initially causing the resources of the first base station to remain held in reserve, and

at a later timepoint determined by a fixed period of time predefined at a beginning of the handover, deleting the link data from the first base station and freeing up the resources of the first base station, the mobile station comprising:

an arrangement for reactivating the link with the first base station if the handover is unsuccessful.

Claim 18 and, by stipulation, claim 16 differ only in ways “immaterial to this appeal.”

The district court had construed the claims to state that the mobile station “achieves a handover” which is then described as a series of steps.  Because the claim described the mobile station both in terms of this method and separately in terms of “an arrangement,” the court therefore invalidated it as indefinite.

The Federal Circuit, however, determined that the “that achieves a handover” clause described “network”, not “mobile station.”  First, “[m]odifiers should be placed next to the words they modify”, and the clause was closer to “network.”  Second, the repetition of “mobile station” in the seventh paragraph “would read in a disjointed manner” if the preceding functions also referred to the mobile station.  And third, the specification confirmed that the base stations of the network, not the mobile station, perform the functions in the handover.

HTC and the district court both tried to overcome this language by pointing to the prosecution history, but the Federal Circuit recalled Phillips v. AWH for the principle that prosecution “often lacks the clarity of the specification and thus is less useful for claim construction purposes.”  Although the applicants had referred to the “claimed process” of claim 1 during prosecution, the Federal Circuit found this insufficient to overcome the “plain language of the claims and the specification” that showed the mobile station was an apparatus.  A single reference to “steps” was similarly unpersuasive.

As the claims did not claim both the apparatus and the process, the Federal Circuit ruled that they did not violate the rule of IPXL.  They were instead merely in a “preamble-within-a-preamble” form, which was “unconventional” but valid under Microprocessor Enhancement Corp. v. Texas Instruments.

General Structure for Means-Plus-Function Claims May Be Implied, Not Explicit

The court then examined HTC’s alternate argument that the claims were indefinite under § 112 ¶ 6.  The parties had agreed that the “arrangement for reactivating” in all three claims was a means-plus-function limitation, which requires disclosure of a corresponding structure in the specification.  But the district court found, and the Federal Circuit agreed, that no literal disclosure was necessary, because a person skilled in the art “would understand that the mobile device would have to contain a processor and transceiver.”  Expert testimony, including the deposition of HTC’s own expert, supported this assessment.

The Federal Circuit also noted that a detailed hardware configuration was not necessary; “reference to such general-purpose processors will suffice” to make the claim definite if the description also discloses an algorithm that “describe[s] a means for achieving [the reactivating] outcome, not merely the outcome itself.”  At the trial level, HTC had argued exclusively that the hardware disclosure was insufficient with or without algorithms, and did not contend in the alternate that the disclosed algorithms were inadequate, even when IPCom argued in briefs that they were.  Therefore, the Federal Circuit did not let HTC raise the argument on appeal, finding no reason not to apply the general appellate policy against considering issues not raised at trial.

With no other cause to invalidate, the panel reversed, declaring all claims valid.

Post-Case Analysis

The Federal Circuit creates little new law here, but reminds us of some existing principles for both prosecutors and litigators.  Prosecutors are reminded of the importance of clarity in drafting.  Although IPCom’s claims survived their second interpretation, clearer attachment of nouns to modifiers would have avoided the initial invalidation.  Simple use of commas or semi-colons to break up the clauses, and then leading each clause with its own element, can be an invaluable rule of thumb.  For instance, opening with “A mobile station for use with a network, the network including…” would have made it undeniable what the following paragraphs referred to.

Similarly, although the courts will treat a structure as implicitly disclosed when it is “understood” by those skilled in the art, it is better to be explicit.  What is “understood” can vary from expert to expert, and had HTC’s own expert not admitted to the validity of IPCom’s position, the patent might well have failed on this point.  Also, this rule has less benefit when the structure is not a processor or similar; that is, one that must be thoroughly defined on its own terms instead of in the context of algorithms.

Means-plus-function claims are especially vulnerable to a few mistakes that IPCom nearly stumbled over.  Prosecutors should always remember to check their claims for IPXL-style indefiniteness, as the drafter is already thinking in terms of the corresponding process and may forget to make the distinction.  Double-checking for the appropriate structure for each function – including the algorithm, if software is involved – is also vital.

Finally, in all cases, proofreading is key.  A second pair of eyes to check for these and other mistakes during drafting can save your client from both a higher risk of invalidity and unnecessary legal fees the first time he tries to enforce.

Litigators, meanwhile, are reminded of the importance of attacking the opponent from every reasonable angle; in this case, to challenge the structure disclosure on both hardware and software.  It is sometimes splitting hairs to say what specific sub-(sub-sub-)issues a party has failed to address, and HTC might have argued that, by bringing up structure disclosure at all, they preserved all aspects of it for appeal.  But at minimum, once IPCom reminded HTC of the algorithm issue in its own briefs, HTC needed to respond.  Leaving that argument unchallenged cost them dearly.

The opinion is available at: http://www.cafc.uscourts.gov/images/stories/opinions-orders/11-1004.pdf

Oral arguments may be heard at: http://oralarguments.cafc.uscourts.gov/default.aspx?fl=2011-1004.mp3

Federal Circuit Says § 101 Requires “Concrete Steps” In Software Methods

By Chris Reaves

Dealertrack, Inc. v. Huber, No. 2009-1566, Fed. Cir. Jan. 20 2012

The Federal Circuit on January 20 once again addressed the question of 35 U.S.C. § 101 eligibility for a software method.   The court found the patent at issue closer to CyberSource than Ultramercial, drawing a distinction between “a practical application with concrete steps” and a mere declaration that the method is “computer-aided.”  It also looked to the interpretation of claims, expanding the rule of Phillips v. AWH to cover exemplary lists.

Background

Plaintiff Dealertrack, Inc. owns three patents related to the same concept: the use of a computer system as a loan intermediary, converting one generalized application for a car loan into several bank-specific applications and processing said applications with their respective banks.  The original patent, U.S. Patent No. 5,878,403 (‘403), was filed in 1995.  Later, 7,181,427 (‘427), patenting the computer system, was filed as a continuation-in-part, and 6,587,841 (‘841), patenting the method, was filed as a divisional application.  All three patents issued, with the latter two claiming priority to ‘403.

Dealertrack sued the defendants, Finance Express, LLC (with its president David Huber) and RouteOne, in the Central District of California, claiming their respective loan management systems infringed on all three patents.  The defendants filed motions for summary judgment on four grounds: that ‘841 was not infringed given its proper interpretation, that certain means-plus-function claims in ‘841 were invalid as indefinite under § 112, that all applicable claims of ‘427 were abstract and ineligible for patent protection under § 101, and that ‘427 did not properly claim priority to ‘403.

The district court found for Dealertrack on indefiniteness and priority but for the defendants on non-infringement and ineligibility, and dismissed the case.  Dealertrack appealed the dismissal to the Federal Circuit and the defendants cross-appealed on the indefiniteness and priority findings.  Judges Linn and Dyk, with Senior Judge Plager, heard the appeal.

Claim Interpretation: Multiple Examples Are Not Always Exhaustive

The preamble of ‘841’s independent claim 7 reads:

“A computer based method of operating a credit application and routing system, the system including a central processor coupled to a communications medium for communicating with remote application entry and display devices, remote credit bureau terminal devices, and remote funding source terminal devices . . .”  (Emphasis added.)

“A communications medium” is also one of the elements of independent claims 12 and 14.  Exemplary mediums are listed in the description as follows:

“Although illustrated as a wide area network, it should be appreciated that the communications medium could take a variety of other forms, for example, a local area network, a satellite communications network, a commercial value added network (VAN) ordinary telephone lines, or private leased lines…  The communications medium used need only provide fast reliable data communication between its users…”

During prosecution of the parent ‘403 patent, the examiner allowed “the Internet” to be added explicitly to a corresponding exemplary list, and ‘841 incorporated ‘403 by reference.  However, in its interpretation of ‘841, the district court interpreted “communications medium” to exclude the Internet, finding its omission in ‘841, in an otherwise thorough list, to be a waiver of its inclusion regardless of ‘403.  Therefore, because both the defendants’ systems used only the Internet for communication, neither had infringed.

The Federal Circuit disagreed, finding that the district court had ascribed too much importance to the list’s contents.  “The disclosure of multiple examples,” it ruled, “does not necessarily mean that such list is exhaustive or that non-enumerated examples should be excluded.”  Rather, it found the case similar to Phillips v. AWH Corp., where a single embodiment listed in a description had not necessarily limited the meaning of the corresponding claims.  Multiple examples did not change the Phillips rule, especially as the context of the ‘841 list – leading with “for example” and closing with “The communication medium used need only provide…” – indicated that it was not meant to be exhaustive.

The district court also found waiver of the term in that the “the Internet” had also been named in ‘841 but the examiner had removed the reference post-allowance.  However, the examiner had given no reason for this removal, and there was no prosecution history suggesting that Dealertrack had deliberately surrendered the term.   The Federal Circuit therefore attached no importance to this removal.  And although the defendants argued that, in 1995, the Internet would not have been seen as “reliable” or “secure” as the description required, the court found they had not presented sufficient evidence on this point.

The court also rejected arguments by the defendants to narrow the meaning of other terms in the patent, noting that explicit embodiments of the claims fell outside the supposed interpretation. It did, however, use the rule of WMS Gaming to apply disclosed software algorithms, not a computer itself, as the structure for claim 12 and 14’s “central processing means” element.  Because claim 14 also stated that the central processing means “further provides for tracking pending credit applications,” without disclosing a tracking algorithm, the court invalidated claim 14 and its dependent claims as indefinite under § 112 ¶ 6.

§ 101: Computer-Aided Methods Require “Practical Application With Concrete Steps”

Claim 1 of ‘427 reads, in total:

1. A computer aided method of managing a credit application, the method comprising the steps of:

[A] receiving credit application data from a remote application entry and display device;

[B] selectively forwarding the credit application data to remote funding source terminal devices;

[C] forwarding funding decision data from at least one of the remote funding source terminal devices to the remote application entry and display device;

[D] wherein the selectively forwarding the credit application data step further comprises:

[D1] sending at least a portion of a credit application to more than one of said remote funding sources substantially at the same time;

[D2] sending at least a portion of a credit application to more than one of said remote funding sources sequentially until a finding [sic, funding] source returns a positive funding decision;

[D3] sending at least a portion of a credit application to a first one of said remote funding sources, and then, after a predetermined time, sending to at least one other remote funding source, until one of the finding [sic, funding] sources returns a positive funding decision or until all funding sources have been exhausted; or,

[D4] sending the credit application from a first remote funding source to a second remote finding [sic, funding] source if the first funding source declines to approve the credit application.

When the district court examined ‘427 for § 101 ineligibility, the Supreme Court had yet to rule in Bilski, and the “machine or transformation” test was still definitive.  Dealertrack had conceded lack of transformation and the district court had found that a “general purpose computer” was not a “particular machine” under the existing test.

The Federal Circuit examined the case under the new standard but came to the same conclusion.  While recognizing that, as in Research Corp., abstractness must “exhibit itself so manifestly as to override the broad statutory categories of eligible subject matter,” the court found that ‘427 crossed this line.  Namely, it claimed “the basic concept of processing information through a clearing-house…”   The court found the “computer aided” limitation in the preamble equally abstract and insufficient to save the patent “without specifying any level of involvement or detail,” and distinguished ‘427 from the patent in Ultramercial, which also described “a practical application with concrete steps requiring an extensive computer interface…”  (Perhaps foolishly, Dealertrack conceded in oral arguments that ‘427 would not be § 101 eligible had “computer aided” been omitted.)

Dealertrack argued that disclosed algorithms in the description made the patent less abstract, but the court noted that the claims themselves were not limited to any algorithm, under an interpretation to which Dealertrack itself had already conceded.  The court also found irrelevant that the method was limited to car loan applications, comparing it to that of Bilski, where a method of hedging was no less abstract for being limited to the energy market.  The court therefore invalidated the applicable ‘427 claims as ineligibly abstract.

On this point alone, Senior Judge Plager dissented.  Without commenting on the thrust of the panel’s argument, he proposed that courts should “not foray into the jurisprudential morass of § 101 unless absolutely necessary” instead looking first to whether they can invalidate under other sections of the Patent Act before the court.  But the majority, while calling Plager’s goal “laudable,” noted that the Supreme Court had called § 101 a “threshold test” in Bilski.  They further noted that no other argument for ‘427’s invalidity was presented in the initial motions for summary judgment; although Dealertrack had argued that ‘427 was clearly valid under § 103, the defendants had been happy to leave that question to the jury.

Because it invalidated ‘427 under § 101, the panel did not address whether the same patent failed to claim priority to ‘403.

Post-Case Analysis: Tell Us Why You Need A Computer

The gray area between eligible and ineligible methods narrows again, albeit slower than anyone would like.  As previously noted on this blog, Ultramercial suggested three applicable factors in the analysis.  This case emphasized two: the software must be specialized and well-described in the claims, and the claim must explain why the software is necessary to the claimed method.  Had the claims in this case restricted their scope to the algorithms in the description, they might have survived the panel’s analysis.  Unfortunately, as the defendants noted in oral arguments, Dealertrack had been arguing the opposite.  Existing patent owners might be wise to concede to a narrower interpretation of their own software patents, and future drafters should include a more elaborate disclosure of the software, algorithms, and computer configurations involved.  Most importantly, the computer should be explicitly necessary, not merely helpful, to the claimed method.

Many commenters (including both Chief Judge Randall Rader and Professor Dennis Crouch of the University of Missouri) have argued, like Plager, that § 101’s jurisprudence is unpredictable and overused, and that public policy should encourage its disuse.  A few have gone farther and proposed that § 101 should be reinterpreted by the courts or rewritten by Congress to allow truly “anything under the sun,” leaving the rest of the Act in place to catch the existing, obvious, or abstract.  To both groups, the majority’s response to Plager will be a serious disappointment, and a step back from the direction suggested in Classen and Ultramercial (albeit in apparent dicta).  If § 101 is indeed a threshold in the sense of “it must be considered,” and if its borders remain vague, defendants will have little reason not to include an ineligibility argument against method claims and hope for the best.  We can hope that the Supreme Court itself will clarify the standard in Mayo Collaborative Services v. Prometheus Labs, but with its past pattern of insisting on a case-by-case analysis, it is more likely to leave the Federal Circuit buried in § 101 cases.

Less vogue a topic but still important is the ruling on ‘841’s interpretation.  Exemplary but non-exhaustive lists of embodiments are common practice, and the court’s opinion protects the resulting patents from unintended limitations.  However, Dealertrack was in part protected by the Internet’s mention in the issuing parent patent; other patents are unlikely to have this extra safety net.  Therefore, a good patent drafter should still make it absolutely clear in context that such lists are indeed non-exhaustive, with phrases such as “for example”, “include but is not limited to”, “many other embodiments and modifications should be apparent”, and “need only [do X] to meet the needs of the invention”.

For software patent prosecutors, this case is also a good reminder of the WMG Gaming rule: means-plus-function claims cannot be satisfied with disclosure of “a computer” or similar.  Only a computer with specific algorithms or software for the job will do.  Any good draft should therefore either avoid means-plus-function language or be as exhaustive as possible in the description.

The opinion is available at: http://www.cafc.uscourts.gov/images/stories/opinions-orders/09-1566.pdf

Oral arguments may be heard at: http://oralarguments.cafc.uscourts.gov/default.aspx?fl=2009-1566.mp3

For Rader’s comments on § 101, see Classen Immunotherapies Inc. v. Biogen IDEC, 100 USPQ2d 1492, 1505-07 (Fed. Cir. 2011).  For Crouch’s, see Operating Efficiently Post-Bilski by Ordering Patent Doctrine Decision-Making, 25 Berkeley Tech. L.J. 1673 (2010).

Federal Circuit Confirms Expansive View of Patent Eligibility

In Ultramercial LLC v. Hulu LLC, 657 F.3d 1323 (Fed. Cir. 2011), Ultramercial LLC’s patent claims a method “for distribution of products over the Internet via a facilitator”. As part of the claim, the facilitating web server offers a “sponsor message” to a potential viewer, who, if he views the sponsor message, is then allowed access to a desired “media product”. The claim also recites limitations related to interactions with the sponsors both before and after delivery of the sponsor message. Continue reading

Federal Circuit Clarified Abstractness in the Context of Software

Court Finds Software Patentable Unless Unmistakably Abstract

In Research Corp. Technologies, Inc. v. Microsoft Corp., Civ. Case. No. 10-1037 (Fed. Cir. December 8, 2010), Research Corporation Technologies, Inc. (“RCT”) owns six related patents: U.S. Patent Nos. 5,111,310 (“’310 patent”); 5,341,228 (“’228 patent”); 5,477,305 (“’305 patent”); 5,543,941 (“’941 patent”); 5,708,518 (“’518 patent”); and 5,726,772 (“’772 patent”).  The six related patents relate to digital image half toning which allows a computer to create more colors than otherwise possible with the limited number of pixel colors available in a typical printer.  In order to measure the quality of the half toning, the inventors created a method of detecting a power spectrum which measures the frequency of the dots used to create the half tone and compares this power spectrum to an ideal spectrum stored in a mask.  By comparing the stored mask and detected power spectrum, the method determines half tone quality. Continue reading

Central District of California Extends Bilski Holding to Find Computer Aided Method Not Tied To A Particular Machine If Merely Tied to Computer

In Dealertrack, Inc. v. Huber et al., CV 06-2335 AG (C. D. Calif. July 7, 2009), the patent at issue, U.S. Patent No. 7,181,427 (the ‘427 patent), is directed toward an automated credit application system.  Dealertrack at 2.  The independent claim at issue, claim 1, recites a “computer aided method” of managing a credit application.  The method consists of steps of receiving credit application data and forwarding the application and subsequent funding decisions to remote devices. Id. at 2.  The only issue in the case was the validity of the claims under 35 U.S.C. § 101.

Since Dealertrack did not argue the “transformation” prong of the test outlined in In re Bilski, 545 F. 943 (Fed. Cir. 2008) (en banc), the court considered only the “particular machine” prong of the Bilski test – whether the claimed process was tied to a particular machine.  Dealertrack at 5.  However, the Bilski decision provided no guidance as to the nature of the “particular machine”, leaving such consideration to future cases.  Id. at 5 (citing In re Bilski, 545 F. 943, 962 (Fed. Cir. 2008)).  The court therefore turned to several recent non-precedential decisions from the Board of Patent Appeals and Interferences, which held that general purpose computers were not “particular machines”.  In Ex Parte Gutta, which was decided before Bilski, the Board held that a process claim “for use in a recommender” was not tied to a particular machine because the body of the claim recited a mathematical algorithm and the only recitation of a machine in the claim was an intended use clause in the preamble.  Ex Parte Gutta, 84 USPQ.2d 1536 (B.P.A.I. 2007).  In Ex Parte Nawathe, the Board found that a claim to a “computerized method” was not tied to a particular machine because the specification discussed only a general-purpose computer.  Ex Parte Nawathe, No. 2007-3360 (B.P.A.I.  Feb. 9, 2009).  Finally, in Ex Parte Cornea-Hasegan, the Board held that a method including steps performed by a “processor” recited merely a general purpose computer.

In light of these Board decisions, the district court found that the ‘427 patent’s claims were not tied to a particular machine.  Although the claims referred to a remote application entry and display device and remote funding source terminal devices, the specification did not indicate how these devices were to be programmed.  Furthermore, during claim construction, the district court interpreted the terms as indicating any device, including a dumb terminal.  Since the devices did not require special programming, and the specification provided no guidance as to the nature of the devices, the district court found that the method was not tied to a particular machine.  The court accordingly ruled that the claims at issue, not being tied to a particular machine or transforming an article into a different state or thing, were invalid under 35 U.S.C. § 101.

Significance for Patent Applicants and Owners

While this case is on appeal, Dealertrack gives further evidence as to the confused state of computer software based inventions due to the Federal Circuit’s en banc opinion In re Bilski, 545 F. 943 (Fed. Cir. 2008).  While the Federal Circuit expressly limited its decision to methods, the United States Patent and Trademark Office and the District Courts are taking opposite approaches as to whether In re Bilski should be extended to all forms of machines and apparatuses to effectively render software unpatentable in any form.  Until the Supreme Court renders its decision In re Bilski, applicants will be unable to determine whether their software-based inventions can be patented, and exactly how much detail needs to included in an application in order to transform the computer into the type of “particular machine” which the District Court in Dealertrack found was required for purposes of determining patent-eligible subject matter under 35 U.S.C. §101.

 

Feature Comment: The Heightened Disclosure Requirement for Software Claimed Using Means-Plus-Function Limitations

By Gregory L. Clinton[1]

I.        INTRODUCTION

The question of whether a computer component discloses adequate structure for a corresponding means-plus-function claim limitation has been a recurrent issue in recent Federal Circuit jurisprudence, most recently in the cases of Blackboard Inc. v. Desire2Learn Inc., 91 U.S.P.Q.2d 1481, 1483 (Fed. Cir. 2008).  and Net MoneyIN Inc. v. VeriSign Inc., 88 U.S.P.Q.2d 1751 (Fed. Cir. 2008).    In both cases the specification of the patents at issue did not provide adequate support for means-plus-function limitations in the claims.  Although the use of means-plus-function claims has fallen out of favor in recent years, means-plus-function claims remain useful in some situations.  Accordingly, a brief discussion on what to include in a specification to ensure proper support for such claims may be helpful to avoid problems during prosecution or litigation

II.       ENABLEMENT OF SOFTWARE MEANS-PLUS-FUNCTION CLAIMS

Both of the two cases previously mentioned, Blackboard and Net MoneyIN, revolved in part around whether the specification provided adequate support for particular mean-plus-function limitations.  In both cases, the specification was found not to provide adequate support because the specifications merely described a general purpose computer, without any corresponding “structure” to which the means-plus-function claims could be tied.  In Blackboard, the patent was directed toward an Internet-based educational support system.  Blackboard Inc. v. Desire2Learn Inc., 91 U.S.P.Q.2d 1481, 1483 (Fed. Cir. 2008).  The claim limitation at issue recited “means for assigning a level of access to and control of each data file based on a user of the system’s predetermined role in a course…”  Id. at 1490.  Blackboard argued that the corresponding structure was an “access control manager” disclosed in the specification, citing to a particular paragraph describing the access control manager in largely functional terms.  Id.

The district court disagreed, finding the specification lacking the necessary structure to support the “means for assigning”, and the Federal Circuit affirmed.  Blackboard at 1490.  The Federal Circuit found that the access control manager was nothing more than a “black box”.  Id.  The specification provided no insight as to how the access control manager carried out the disclosed functionality.  Id.  In particular, the specification did not describe the structure of the access control manager or explain the process by which the access control manager performed the disclosed function.  Id.

The Federal Circuit rejected Blackboard’s arguments that the disclosed access control manager could be any computer device or program that performs the function of access control.  Blackboard at 1491.  The Federal Circuit viewed the argument as an attempt to avoid the requirement to disclose the structure associated with the means-plus-function limitation.  Blackboard was attempting to avoid the “price” of the means-plus-function limitation (namely, the requirement to disclose the corresponding structure) by attempting to claim the structure in purely functional terms, without any limit imposed by a structure disclosed in the specification.  Id.  What is important, the Federal Circuit held, was the means for achieving the intended outcome, not a mere description of that outcome.  Id.

One of the cases cited in the Blackboard decision, Net MoneyIn, Inc.  v. Verisign, Inc., expanded on the requirements in the disclosure for means-plus-function claims and computer-related structures.  The limitation in that case was a “means for generating an authenticating indicia” as part of a recited “first bank computer”.  Net MoneyIn, Inc. v. Verisign, Inc., 88 U.S.P.Q.2d 1751, 1755 (Fed. Cir. 2008).  After concluding that the limitation was properly construed as a means-plus-function limitation, the Federal Circuit looked to the specification for the corresponding structure, and found only a description of a general purpose bank computer that could be programmed to perform the recited function.  Id

The Federal Circuit reviewed prior holdings that the structure disclosed in the specification must be more than simply a general purpose computer.  Net MoneyIN at 1755.  If the disclosed structure is a computer, then the disclosure must be directed not to the general purpose computer itself, but to the algorithm which the general purpose computer is programmed to perform.  Id.  The operation of the algorithm performed by the general purpose computer transforms the general purpose computer into a special purpose computer designed to perform the disclosed algorithm.  Id.  Thus, if the specification fails to disclose the algorithm performed by the general purpose computer to carry out the recited function, then the claim is invalid under 35 U.S.C. § 112 for failing to disclose the necessary structure corresponding to the means-plus-function limitation.  Id. at 1757.

A.  GENERAL PURPOSE COMPUTER PERFORMING A DISCLOSED ALGORITHM

The Federal Circuit’s decision in the Blackboard and Net MoneyIn cases relied on its prior decision in WMS Game Inc. v. Int’l Game Tech., which in turn drew upon the landmark State Street Bank and Alappat decisions.  The patent at issue in WMS Gaming was directed toward an improved slot machine.  WMS Gaming, Inc. v. Int’l Game Tech., 51 U.S.P.Q.2d 1385, 1387 (Fed. Cir. 1999).  In traditional mechanical slot machines, reducing the probability of winning required either increasing the number of reels in the machine or increasing the number positions in each reel, neither of which appealed to players.  Id. The patent at issue in the case disclosed an electronically controlled slot machine.  Id.  The reels in the machine served only to display the results selected by the electronic control system.  Id.  The control system randomly selects a number from a range greater than the number of positions on the reel, looks up the corresponding result in a mapping table, and controls the reels to display the selected result.  Id.

The claim term at issue recited “a means for assigning a plurality of numbers representing said angular positions of said reel, said plurality of numbers exceeding said predetermined number of radial positions such that some rotational positions are expressed by a plurality of numbers.”  The district court interpreted the term to refer to “any table, formula, or algorithm for determining correspondence between the randomly selected numbers and rotational positions of the reel.”  WMS Gaming at 1391.  The Federal Circuit, however, found this interpretation to be too broad.

The flaw in the district court’s construction, the Federal Circuit held, lay in the district court’s failure to limit the claim to the algorithm disclosed in the specification.  WMS Gaming at 13.  The Federal Circuit reasserted its opinion in Alappat that a general purpose computer programmed to perform a given algorithm in effect becomes a new computer.  The general purpose computer in effect becomes a special purpose computer performing the disclosed algorithm.  Id. (citing In re Alappat, 31 U.S.P.Q.2d 1545, 1558 (Fed. Cir. 1994)).  Thus, in the context of means-plus-function claims, the structure is not the general-purpose computer, but rather the special purpose computer programmed to carry out the disclosed algorithm.  Applying this construction to the “means for assigning” limitation, the Federal Circuit held the corresponding structure to be a processor programmed to carry out the algorithm disclosed in Figure 6 of the patent.

The algorithm was disclosed in the patent at issue in WMS Gaming in the form of a diagram illustrating the operation of the algorithm.  However, patent applicants can express the algorithm in a variety of ways.  The algorithm can be explained as a diagram (as in WMS Gaming), as a mathematical formula, as a flowchart, or in prose. See Finisar Corp. v. DirecTV Group, Inc., 86 U.S.P.Q.2d 1609, 1623.

B.  SPECIAL PURPOSE COMPUTER

The Federal Circuit’s justification for defining algorithms as the important element for computer-related structures corresponding to a means-plus-function limitation relies upon the concept that a general purpose computer performing a disclosed algorithm can be seen as a special-purpose computer programmed to perform the function.  From this premise, it follows that a special-purpose computer could also be used as a sufficient structure.  Instead of disclosing an algorithm and a general-purpose computer to perform the algorithm, a patent applicant can instead disclose a special-purpose computer as the corresponding structure for a means-plus-function limitation.  Such a disclosure could come in the form of, for example, a detailed circuit diagram or schematic of the individual components of the special purpose computer.  A patent applicant may find such an alternative useful in situations where the novelty or inventive aspect lies more in the structure of the component than in the algorithm to be performed.

III.  EXAMPLES

Federal Circuit jurisprudence permits the algorithm to be disclosed in a variety of ways.  The algorithm may be disclosed by a flowchart, a diagram, a formula, or prose.  Finisar at 1623.  Several cases illustrate the variety of ways in which a patent can disclose the necessary algorithm to support a means-plus-function case.  For example, the algorithm in WMS Gaming was disclosed by a combination of prose description in the specification and a figure.  See U.S. Patent No. 4,448,419 at Fig. 6 and col. 4, line 55 – col. 5, line 5.

In Harris Corp. v. Ericsson Inc., the algorithm was disclosed in a flowchart and corresponding textual description.  Harris Corp. v. Ericsson Inc., 75 U.S.P.Q.2d 1705, 1714 (Fed. Cir. 2005) (citing to various portions of U.S. Patent No. 4,365,338, including a flowchart shown in Figs. 8A and 8B).  The Federal Circuit disagreed with the district court’s finding that the recited “time domain processing means” corresponded to the disclosed “symbol processor”, because the symbol processor did not incorporate a disclosed algorithm.  The Federal Circuit instead found the disclosed structure to correspond not only to the symbol processor (which in turn included a support processor and a fast array processor) but also to a data recovery algorithm performed by the support processor.  Id.  The data recovery algorithm was disclosed in the specification as a flowchart diagram and corresponding description.  SeeU.S. Patent No. 4,365,338 at Figs. 8A, 8B, and 9.

The knowledge of a person of ordinary skill in the art can also support a disclosed structure or algorithm.  The patent at issue in AllVoice Computer PLC v. Nuance Communications Inc. recited an “output means” as a part of a system to facilitate speech recognition.  AllVoice Computer PLC v. Nuance Communications Inc., 84 U.S.P.Q.2d 1886, 1891.  AllVoice’s patent was directed toward an interface that converted spoken words into a data format compatible for word processors.  Id. The specification disclosed a speech recognition interface which, as part of its functionality, received a translated word and output the word via a dynamic data exchange (DDE) protocol.  Id.  Expert testimony established that this protocol was well known to persons of skill in the art of Windows programming.  Id. Thus, the person of ordinary skill in the art would have been well apprised of a structure corresponding to the disclosed DDE protocol. 

The AllVoice decision contrasts with the Blackboard and Net MoneyIn cases discussed above.  In both cases, the patentee attempted to rely upon the knowledge of a person of ordinary skill in the art to support the disclosed structure, but of the three, only AllVoice succeeded in convincing the Federal Circuit.  The distinguishing feature lies in the  extent to which the parties relied upon the knowledge of the person of ordinary skill in the art.  The disclosures of the Net MoneyIN patent and the Blackboard patent contained no element or internal structure of the various components.  The components were simply “black boxes” that a person of ordinary skill in the art could have programmed to perform the recited function.  Compare Net MoneyIN, 88 U.S.P.Q. 2d at 1756 and Blackboard, 91 U.S.P.Q.2d at 1491.  However, the disclosure of the AllVoice patent described the structure of the component in terms that would be understood by a person of ordinary skill – namely, the DDE component, which was a structure known in the art for carrying out a particular function.  This provided enough structure to render the corresponding means limitation definite.

IV.  CONCLUSION

As can be seen in the above cases, patent application drafters should pay special attention to the specification if means-plus-function claims are to be used.  If computer components are to be disclosed as the structure corresponding to a means-plus-function, either the specific structure of the corresponding component(s) should be disclosed (i.e. a special-purpose computer), or the algorithm carried out by the corresponding component(s) should be disclosed.  The algorithm can be disclosed in a variety of ways, including flowcharts, written explanation, diagrams, mathematical formulae, or a combination of these.  By disclosing the algorithm and the particular components, the careful patent drafter can avoid the definiteness problems that arose in Net MoneyIN and Blackboard.


[1] Gregory L. Clinton is an associate at Stein McEwen, LLP.  The opinions in this article do not represent the official positions of Stein McEwen, LLP.

Federal Circuit Invalidates Software Claim for Indefiniteness

Federal Circuit Finds Means plus Function Element in Claim Indefinite Since Specification Did Not Describe Structure in Specification

In Blackboard, Inc. v. Desire2Learn Inc., 91 U.S.P.Q.2D 1481 (Fed. Cir. 2009), the Court of Appeals for the Federal Circuit (CAFC) deliberated on two issues raised by the respective parties on appeal from the United States District Court for the Eastern District of Texas.  Specifically, Desire2Learn contested the district court’s holding of infringement of claims 36-38 of Blackboard’s U.S. Patent No. 6,988,138, which relates to an Internet-based educational support system and corresponding methods.  In turn, Blackboard cross-appealed the District Court’s ruling, on summary judgment, that claims 1-35 of the ‘138 patent are indefinite.

Single login feature

Turning to the first issue, Desire2Learn argued that claims 36-38 do not include a “single login” limitation for the recited method of providing an online education system, and thus were anticipated by two prior art references.  The ‘138 patent discloses a system that allows for online management of information relating to individual courses.  Of particular benefit is the ability to allow a user to use a single login to access multiple courses with multiple respective roles in those courses (i.e., the “single login” limitation).  For example, a graduate student who is a student in a first course and a teacher in a second course could use a single login to access both courses and access materials for both courses according to the corresponding role of the graduate student in the respective course. 

Blackboard asserted that the “single login” feature is its patent’s essential improvement over prior art systems, and is recited in every claim including independent claim 36 (from which claims 37 and 38 depend).

Claim 36 recites:

An [sic] method for providing online education… comprising the steps of:

establishing that each user is capable of having redefined [sic: “predefined”] characteristics indicative of multiple predetermined roles in the system and each role providing a level of access to and control of a plurality of course files;

providing a predetermined level of access and control over the network to the course files to users with an established role as a student user enrolled in the course; and

providing a predetermined level of access and control over the network to the course files to users with an established role other than a student user enrolled in the course.

Though claim 36 does not explicitly provide for a single login, Blackboard argued that such a single login is required because the term “user” refers to an electronic user account that is defined by a single user name and password combination.  Thus, the user being capable of having multiple roles is an electronic user having a single name and password combination, i.e., a single login. 

For support, Blackboard pointed to the specification, which sometimes uses the word “user” to ostensibly refer to an electronic representation of a person.  However, the Federal Circuit (Judges Bryson and Moore, andCudahyby designation) disagreed.  Specifically, the Federal Circuit reasoned that the specification typically uses the term “user account” to refer to the electronic representation of the user.  Moreover, aside from a few shorthand references, the specification also typically uses the term “user” to refer to a flesh-and-blood person.  Also, of the four embodiments described in the specification, only one includes the single login feature.  Therefore, the Federal Circuit found that specification does not require that the recited “user” necessarily refers to an electronic user in the context of a single login.

The Federal Circuit further reviewed other claims and noted the relationship between claim 1 of the ‘138 patent and claims 24 and 25 which depend from claim 1.  While claim 1 includes the same pertinent language as claim 36, dependent claims 24 and 25 recite the single login limitation.  Thus, the Federal Circuit reasoned that as claim 1 cannot be construed to require a single login without making claim 25 redundant, claim 36, too, does not require the single login.  Therefore, in its decision, the Federal Circuit affirmed that a dependent claim adding a particular limitation gives rise to the presumption that the limitation is not present in the independent claim.

As a result, the Federal Circuit held that, as there was not recited single login feature, claims 36-38 were invalid for anticipation by prior art systems. 

Indefiniteness

Turning now to the 35 U.S.C. 112 issue, Blackboard challenged the district court’s invalidation of claims 1-35 as indefinite.  Referring to independent claim 1 as representative (claims 2-35 depend from claim 1), claim 1 includes four means-plus-function clauses, including “means for assigning a level of access to and control of each data file based on a user of the system’s predetermined role in a course.”  As the limitation is written in means-plus-function form, it is construed to cover only the structure described in the specification and equivalents thereof, as prescribed by 35 U.S.C. 112, paragraph 6.

Blackboard argued that the structure performing the assigning function is a software feature known as the “access control manager” (ACM), which is described in only one paragraph of the specification:

Access control manager 151 creates an access control list (ACL) for one or more subsystems in response to a request from a subsystem to have its resources protected through adherence to an ACL.  Education support system 100 provides multiple levels of access restrictions to enable different types of users to effectively interact with the system (e.g. access web pages, upload or download files, view grade information) while preserving confidentiality of information.

The Federal Circuit upheld the district court’s decision that claims 1-35 are indefinite, because the specification failed to sufficiently describe the means by which the ACM creates an ACL.  In particular, the Federal Circuit decided that the lone paragraph describing the ACM does not describe a structure, but merely puts forth an abstraction defined only by a function.  That is, “[t]he specification contains no description of the structure or the process that the access control manager uses to perform the ‘assigning’ function.  Nor has Blackboard ever suggested that the ‘access control manager’ represents a particular structure defined other than as any structure that performs the recited function.”

Blackboard argued that reference to a general-purpose computer satisfies the requisite structural disclosure for the means-plus-function limitation.  Citing Aristocrat Techs. Austl. PTY Ltd. V. Int’l Game Tech., 521 F.3d 1328, 1334 (Fed. Cir. 2008), the Federal Circuit countered that relying on a general structure is equivalent to saying “that the function is performed by a computer that is capable of performing the function.”  Moreover, citing Net MoneyIN, Inc. v. VeriSign, Inc., 545 F.3d 1359, 1367 (Fed. Cir. 2008), the Court further noted that “when a computer is referenced as support for a function in a means-plus-function claim, there must be some explanation of how the computer performs the claimed function.”  In other words, the Federal Circuit concluded that the specification only describes the function to be performed, but says nothing about how the ACM performs the function (i.e., how the ACM creates the ACL).

Blackboard also argued that the process of creating an ACL through software is well known to one of ordinary skill in the art.  The Federal Circuit was unmoved by this argument, emphasizing that the question before it is not the ability of one of ordinary skill in the art, but whether the specification sufficiently describes a corresponding structure to a claimed means.  In particular, the Court asserted that a patentee cannot avoid sufficiently disclosing a structure because someone of ordinary skill in the art could come up with a means to perform a claimed function.  “That ordinarily skilled artisans could carry out the recited function in a variety of ways is precisely why claims written in ‘means-plus-function’ form must disclose the particular structure that is used to perform the recited function.”

Thus, in holding claims 1-35 to be indefinite, the Federal Circuit concluded that Blackboard is attempting to capture any possible means for performing the assigning function, in violation of 35 U.S.C. 112, by disclosing the means only as a general structure to perform the function.

Significance for Patent Applicants

As set forth below in the feature comment The Heightened Disclosure Requirement for Software Claimed Using Means-Plus-Function Limitations, when describing an invention which is realized using software, it is important that the specification provide a heightened level of detail.  While there is no brightline test, this detail must be more than merely an indication that the particular module can be implemented using a general purpose computer.  Instead, the patent applicant needs to provide details as to how the module operates or is constructed, possibly through the use of flowcharts.  While seemingly limited to limitations relying on 35 U.S.C. 112 ¶6, it is possible that this logic could be extended to other types of functionally-described elements.  Such a potential would appear to be higher during prosecution where the Examiner is entitled to take a broader interpretation of the claims, as demonstrated by the Board of Patent Appeals and Interferences in Ex parte Kenichi Miyazaki, 89 USPQ2d 1207 (BPAI 2008).   Thus, while it is unlikely that levels of detail such as copies of source code would be required, when drafting applications which are being implemented using software, applicants need to ensure that the specification includes structural detail as to the implementation of the software modules.

Federal Circuit Revises Comiskey Decision in light of Bilski

In En Banc Ruling, Federal Circuit Orders New Decision to Address Whether Arbitration Methods and System Comply with Machine-or-Transformation Test  

In In re Comiskey, 89 USPQ2d 1641 (Fed. Cir. 2009) (en banc), the Court of Appeals for the Federal Circuit, sitting en banc, issued an order allowing the merits panel to issue a revised opinion found at In re Comiskey, 89 USPQ2d 1655 (Fed. Cir. 2009) and vacating the prior opinion found at In re Comiskey, 499 F.3d 1365; 84 USPQ2d 1670 (Fed. Cir. 2007).   The revised decision removes the portion of the opinion discussing claims 17 and 42, which were directed to a “system”.  Previously, the court had ruled that these claims satisfied the patentability test of 35 U.S.C. § 101, but were prima facie obvious under 35 U.S.C. § 103(a).  The revised decision instead remands the case to the Board of Patent Appeals and Interferences to determine, in the first instance and in light of the Federal Circuit’s recent decision in In re Bilski, 88 U.S.P.Q.2d 1385 (Fed. Cir. 2008) (en banc), whether these claims are patent-eligible under 35 U.S.C. § 101 but did not include the language arguably making such claims prima facie obvious.

By way of background, Stephen Comiskey filed an application for a method and system of mandatory arbitration.  Claim 1, a method claim, includes steps of enrolling a person in a mandatory arbitration system, incorporating arbitration language into a unilateral contract, requiring the complainant to submit a request for arbitration, conducting and providing support for the arbitration, and determining an award or decision in the arbitration.  Claims 17 and 42 recite a system having modules to carry out these steps; and depending claims 15, 30, 44, and 58 depend from the system or method and recite that “access to the mandatory arbitration is established through the Internet, intranet, World Wide Web, software applications, telephone, television, cable, video [or radio], magnetic, electronic communication, or other communication means.”

The Examiner rejected all claims as obvious under 35 U.S.C. § 103(a).  Throughout the prosecution of the case, the Examiner did not reject the claims on any other grounds.  Comiskey appealed the Examiner’s rejection to the Board, and the Board affirmed the obviousness rejections.  Comiskey appealed to the Federal Circuit.

The Federal Circuit did not reach the Board’s obviousness determination.  Rather, the Federal Circuit found that the claims were “barred at the threshold by [35 U.S.C.] § 101″.  In re Comiskey, 89 USPQ2d at 1659.  The Federal Circuit held that Comiskey’s method claims were not patentable subject matter under 35 U.S.C. § 101, and, in the revised portion of the opinion, remanded to the Board to determine the patentability of the system claims and the method claims explicitly reciting the use of a computer. 

In reaching the decision regarding the method claims, the Federal Circuit first reviewed Supreme Court case law regarding algorithms and business methods.  Citing a long line of Supreme Court precedent, the Federal Circuit concluded that claims reciting an abstract idea per se are not patentable.  However, a claim directed to an abstract idea may be patentable if the claim recites a practical application of the abstract idea.  Such a claim is statutory only if, as employed in the process, the idea is embodied in, operates on, transforms, or otherwise involves a class of statutory 

 

subject matter.  Thus, a mental process standing alone, untied to any other category of statutory subject matter, does not recite patentable subject matter.

The Federal Circuit then turned to Comiskey’s method claims, finding that the exceptions to the “abstract idea” rule did not apply to these claims.  The claims did not require a machine, as Comiskey conceded.  Nor did the claims operate on or transform another class of statutory subject matter: they were not “a process of manufacture” or “a process for the alteration of a composition of matter”.  The claims merely described the steps of arbitration by a human arbitrator.  As such, the claims recited only mental steps, and were therefore not patentable.      The Court treated the system claims separately.  Unlike the independent method claims, the system claims recited limitations that, in their broadest reasonable interpretation, could be interpreted as requiring the use of a machine.  Claim 17 recited, among other limitations, “a registration module”, “an arbitration module”, and “a means for selecting an arbitrator from an arbitrator database.”  The system claims would thus be patentable, since they would be “embodied in” another class of statutory subject matter: machines.  The Federal Circuit did not, however, rule on this point.  Instead, the Federal Circuit remanded the case to the Board to determine, in the first instance, whether the system claims, as well as certain dependent method claims directed toward establishing access to arbitration over the Internet or other media, were patentable under 35 U.S.C. § 101. 

The difference between the 2007 opinion and the revised 2009 opinion lies in the treatment of the system claims.  In the original 2007 opinion, the Federal Circuit in fact ruled that the system claims recited patentable subject matter because, under the broadest reasonable interpretation, the claims could be directed to a machine.  In re Comiskey, 84 U.S.P.Q.2d 1670, 1680 (Fed. Cir. 2007)(withdrawn).  The Federal Circuit remanded these claims to the Board not to determine their patentability under 35 U.S.C. § 101, but to determine whether the mere addition of general purpose computers was obvious under 35 U.S.C. § 103(a).  In the revised opinion, the Federal Circuit retreated from its prior holding of patentability, instead remanding the case to the Board to determine the § 101 issue in the first instance, without reaching the obviousness issue.

The Federal Circuit justified the limited remand to the Board with respect to the system claims in part because, had the Board relied on 35 U.S.C.  § 101 to reject the new claims, Comiskey would have had an opportunity to amend the claims under 37 C.F.R. § 41.50(b).  The different treatment of the system claims and the method claims is curious in that Comiskey would have had the same opportunity to amend the method claims had the Board rejected those under 35 U.S.C. § 101.  The panel may have felt that the method claims presented a clear case of unpatentable subject matter, while the system claims did not.  This position would be consistent with the Federal Circuit’s subsequent In re Bilski decision.  88 U.S.P.Q.2d 1385 (Fed. Cir. 2008) (en banc).  In In re Bilski, the Federal Circuit limited the issue to “what the term ‘process’ means”, subsequently ruling the method claims at issue to be unpatentable subject matter because the method did not transform an article to a different state or thing or was not tied to a particular machine.  Id. at 1389.

Significance for Patent Owners

The Comiskey decision joins the Bilski decision as part of the Federal Circuit’s attempt to define the limits of patentable processes.  The original 2007 decision pointed the way to Bilski, and the 2009 revised decision stands as an application of the Bilski principles.  The Supreme Court, however, may soon have an opportunity to return to this area of the law for the first time since Diamond v. Diehr; Bilski recently filed a petition for certiorari with the Supreme Court.  Unless the Supreme Court weighs in, the principles of Bilski and Comiskey remain the law: a method cannot be patentable unless the method transforms an article to a different state or thing, or is tied to a particular machine.  Purely mental processes, such as Comiskey’s arbitration method, are barred by 35 U.S.C. § 101.  As such, applicants should ensure that methods, both in the specification and claims, specifically recite the processors and machines with which they interact in order to better assure compliance with the Federal Circuit’s new machine or transformation test.