Our Product is Lossy and Slow — Therefore we Don’t Infringe

Iridescent Networks, Inc. v. AT&T Mobility, LLC  (Fed. Cir. 2019)

AT&T’s legal arguments get a win — by proving that its products a loss.

Iridescent’s patent covers a method of providing “high quality” bandwith “on demand” using custom routes (rather than ad hoc packet-by-packet routing) to maximize the availability of bandwidth, minimize packet loss, and reduce latency.

On appeal, the main issue was the meaning of “high quality” network service as used in the claims.  AT&T argued particularly that the term should be defined as having a connection speed of >= 1 mbps; packet loss <= 0.0001%; and latency of < 1 second. Under that definition, everyone agreed that AT&T’s service was not “high quality.”

On appeal, the Federal Circuit agreed.  Although the patent does not define “high quality,” it did provide a chart that appeared to delineate between some services that are high quality and those that are not.

When looking at the prosecution history, this appears to be an easy case. In response to an enablement rejection over the “high quality” term, the patentee explained the term’s meaning has being tied to particular numbers:

High QoS (quality of service) may be viewed in the present application as having speeds varying from approximately 1–300 megabits per second, packet loss requirements that are typically about 10-5, and latency requirements that are typically less than one second. These are commonly used parameters . . . .

Ordinarily, the Federal Circuit looks to the prosecution history when there is some disavowal of scope.  Here, the statement might not have risen to disavowal, but were still considered important because the term “high quality” service did not have a clear ordinary and custom meaning in the field.

In these circumstances, where there is no clear ordinary and customary meaning of a coined term of degree, we may look to the prosecution history for guidance without having to first find a clear and unmistakable disavowal.

Thus, the decision here provides some caution and guidance to patent prosecutors using “coined terms:” (1) even if your term uses “ordinary words” it can be considered coined if it lacks an “ordinary and customary meaning” in the art; (2) the court is more willing to look to prosecution history to understand the meaning of a coined term.

24 thoughts on “Our Product is Lossy and Slow — Therefore we Don’t Infringe

  1. 5

    The right result, for the right reasons. Anytime these hand-waving undefined junk terms are invoked, they should be construed against the patentee (whether for the purpose of finding non-infringement or invalidity).

    Sanctions would have been appropriate here. Did everyone catch the patentee’s absurd “argument” that the term “high quality” shouldn’t be tied to any number?

    1. 5.1

      What’s wrong with an argued definition of a claim term of “high quality” that is not tied to any numbers? Obviously, it was intended to mean high enough to avoid prior art and low enough to be infringed.

      1. 5.1.1

        A problem that I have with such a term is the unbounded side and enablement/possession.

        How “high” is “high quality?”

        Does it cover packet loss to 10 to the minus 15? to the minus 100?

        Latency less than a millionth of a pico second? Down to a Planck time?


          You realize, of course, that 99% of “do it on a computer” claims are similarly unlimited in terms of the maximum rate at which the logic processing occurs. The claims aren’t even limited to existing operating systems. They cover data processing logic in all machine-executable contexts, whether the machine is a prior art data processing machine, or an as-yet-to-be-invented machine that carries out data processing using presently undisclosed (and/or unconceived) structures.

          In other words, the “technology” that is being claimed is … logic. And claiming logic was something that was never permitted in the patent system until very recently, when the door was opened by a handful of activist judges, at least one of whom (a Republican appointee, shockers) later resigned in disgrace. And now we have a bunch of Republican/glibertarians types pushing to expand the system to include claims that protect correlations and other facts (as long as the claim additionally includes some “non-human” step — golly, what a hurdle).


            You realize, of course, that 99% of “do it on a computer” claims are similarly NOT claimed in this “unlimited” drawback manner.

            You are far too eager t jump based on your emotions.

            Engage your reasoning (if any at all is available) first.


              …and then tell me how that project of copyrighting “just logic” is coming along.

              (You lame canards do not get any better just because you mindlessly and endlessly repeat them — they remain flawed at their core)


                tell me how that project of copyrighting “just logic” is coming along.

                Setting aside the incredible d-u-m-b-n-e-s-s of this “comeback” of yours (it’s not even clear whether to call it a non-sequitur or a strawman), I note the implication within the question that copyrighting “just logic” seems for you to represent a problematic endeavor. Am I mistaken about that? Or do you think that an expanded copyright law that would allow someone to protect “just logic” (or “correlations”) would be free of Constitutional defects?

                You’re a very serious person. I’m sure you’ve thought very deeply about this already so your cogent and insightful answer will be posted here shortly. Thank you in advance for the comment.

                1. The comeback is neither a strawman nor a non-sequitur, being as it is directly on point to your canard of attempting to treat software merely as “just logic,” which of course would rule out the ability to earn copyright.

                  But you already knew that.

                  And if you want to make a First Amendment argument, I look forward to you actually making one (mouthing ‘First Amendment’ does not count)

  2. 4

    For some reason my comment doesn’t appear. Will this attempt work better? Could it be the three embedded links?


    Under that definition, everyone agreed that AT&T’s service was not “high quality.”

    I can’t wait to see what US late night show hosts, in particular John Oliver, could do with this admission. 🙂

    Claim 1 provokes uncontrollable yawning (typical telecom patent verbiage with claims reciting boxes talking to other boxes), but the collision between seemingly random claim wording in the drafting with random objections, and the attention given to the examination estoppel, is always interesting from an EP perspective. Looking at the amendments the applicant made I think I would have wielded with wild abandon the infamous EPC added subject-matter club, but this wasn’t at all an issue here.

    US8036119 was issued in the inventor’s name, and the USPTO assignment database records no transfer of title, yet the decision states as a matter of fact : “Iridescent Networks, Inc. (“Iridescent”) is the assignee of U.S. Patent No. 8,036,119 (“the ’119 patent”)”, without any further discussion. Odd. What am I missing?

    1. 4.1

      Re your pt. 2: the parent application to which US 8036119 claims priority has assignments filed for the transfer. Per USPTO rules, those assignments are effective for continuations and divisional applications.

      Though as good practice, you should re-file the assignment against the child application because it makes it easier to locate and confirm ownership.

      1. 4.1.1


        Nice post. Also of note is that the espacenet links also track the lineage to the parent (of this continuation), and at that parent provides the legal status of assignment, so one is on notice that a family assignment is present (without venturing into the USPTO systems).

        Of course, your tidbit is crucial to understanding what the espacenet system provides.

    2. 4.2

      RTF, in the U.S. if ownership of a patent is asserted in the patent suit complaint, and is not contested, the failure to record the assignment in the U.S. patent assignment database is no impediment. The limited legal advantages 35 USC 261 provides have to be weighed against what may be considered an advantage in reducing competitive analysis.

    3. 4.3


      think I would have wielded with wild abandon the infamous EPC added subject-matter club, but this wasn’t at all an issue here.

      Most often when I see a comment like this, it is coming from a euro viewpoint, and misunderstands a number of US points that may differ from that euro perspective.

      First, for some reason, euro’s tend to have a mistaken notion that the US does not have a “no new matter added” rule.

      We do.

      Second, that notion often comes about because euro’s also tend to import their more severe ipsis verbis notion of wanting “exact wording” in order to meet the “no new matter” requirement.

      That is simply not the requirement in the US Sovereign.

      1. 4.3.1

        Of course you in the USA have a prohibition on adding disclosure during prosecution. Every jurisdiction does, the same as every jurisdiction requires novelty, non-obviousness and sufficiency of description at the date of the claim.

        Everybody knows this, including “euro’s”.

        What differs as between jurisdictions is what constitutes (to the deemed reader) the “disclosure” of any particular document. In Europe, for example, it is defined as all that which is “derivable” from the document, directly and unambiguously, but no more than that. Hence, not only the EPO’s strict policing of the admission of amendments in prosecution, but also its corollary, its frequent recognition of novelty over the disclosure of any given prior publication.

        Similarly, every jurisdiction has a Doctrine of Equivalents but jurisdictions differ in their methodology to determine what is “equivalent”. Discussion between jurisdictions is what exposes deficiencies and fallacies of analysis in these matters by the courts in each individual jurisdiction. Judges in the main enjoy such discussions. They provide an opportunity to joust, and exhibit which judges have the sharpest minds.


          And which commentators do not (as in here, your propensity to ignore the fact that different Sovereigns have chosen differently — even (especially) under such things as obviousness (as in, a consistent application of the legal fiction known as PHOSITA).

          But we both know that already, eh?

  3. 3

    Despite the low quality of the judges (basically purchased and unqualified), this looks like the right outcome. This is a good example of how you should be careful using terms that may end up with a definition strictly from your specification.

    It is interesting the tension between this and “functional” terms. The reason that patent attorneys want functional terms is illustrated by this case.

    1. 3.1

      What ALL patent attorneys merely WANT is to GRIFT, don’cha know that by now Night Writer?


  4. 2

    This seems to be an easy decision as noted in the comments. How many times in an Office Action response have we cited to MPEP 2111.01(IV) where an Applicant may be their own lexicographer? I think it’s safe to say that most patent practitioners think long and hard about citing this because of the prosecution history it creates. We know that it will stick with the file if it ever ends up in litigation.

    1. 2.1

      Good point — creation of prosecution history is (and should be) always a consideration during prosecution.

  5. 1

    For me, the curious thing is why the patent owner was so dogged, pursuing its hopeless case all the way to a written decision out of the Federal Circuit.

    Is it perhaps that there is no financial risk in adopting that strategy; that, who knows, keeping the litigation pending indefinitely might lead to a lucrative settlement agreement, especially if the Defendant is embarrassed by the ongoing litigation.

    On this occasion, Defendant was big enough to shrug. But what if it had been a smaller Defendant? How does one best ensure “equality of arms” in civil litigation?

    1. 1.1

      MaxDrei raises an interesting question. However, the purported answer sounding in harassment and embarrassment are — as any attorney would know — unethical reasons.

      With a cavalcade of other possible reasons that entirely fall within the ethical sphere, I have to wonder then why unethical reasons would be proffered.

      The ensuring of “equality of arms” is already present in the US system by way of the ethical code that bonds attorneys on all sides. Perhaps the better question then would be how to provide a more clear or powerful bite to ethical transactions (which, by the by, include unauthorized practice of law by representing others in front of the USPTO without a current registration number).

      Secondarily, I have to wonder about the phrasing in the article of: “Ordinarily, the Federal Circuit looks to the prosecution history when there is some disavowal of scope.

      The phrasing appears to mislead with the suggestion that the courts here looking at the prosecution history is somehow not an ordinary thing, when in fact this type of situation is merely ALSO an ordinary thing.

    2. 1.2

      Re: “How does one best ensure “equality of arms” in civil litigation?” Good question.
      From what is available now, it would be an award of attorney fees to the party being forced to proceed in a “hopeless case,” but only if that was clearly the case – an “exceptional case” under 35 USC 285.
      Of course if the hopelessness was clear from the complaint itself [not the case here] there are other possible fixes.

      1. 1.2.1

        Agreed – mechanisms are already in place.

        The “take” I obtained from MaxDrei’s post was the echo of the “Oh N0 es, Tr011s” mantra.

Comments are closed.