Cisco Sued over OFDM Wireless Standards

Agent Green writes "It's definitely not the first time someone has been sued over a standard, but Wi-LAN is in the process of taking Cisco to court over the OFDM encoding which it claims to have patents for - the standards in question apply to 802.11a/g. Interestingly, this case is being brought in Canada, where the defense needs to prove its case. Might be time to join and expand the patent busting brigade?"

Comments

Trouble

[sigaar]

Patents just cause trouble. See? See?

exactly!

[curator_thew]

Who cares that it may have cost millions of dollars of risk and investment to devise, refine and perfect OFDM and the related technologies ... we just want them to be free for all of us to use, so we definitely should bust their patents.

By the way, I know that it took you years of hard work to earn a salary to pay off your mortgage, but I actually think your house would make for a good party zone, so me and the boys will be around next Saturday night.

Re:exactly!

[kfg]

I actually think your house would make for a good party zone, so me and the boys will be around next Saturday night.

No thank you, but if you'd like to build yourself a house just like it I'd have no particular objections.

KFG

Re:exactly!

[sigaar]

OK, fair enough. I'm throwhing a party on Saturday night. You and the boys are invited to party with us. But you have to bring your own booz, own snacs, own smokes, own women. You're not allowed to come in my house, in fact, stay out of my yard - the sidewalk is as good a pary place as any. You're not allowed to mingle with the rest of the guests. If you do I'll sue you. Let's look at the pattern. And this is in general terms, not just this case. A company figures someting out, patents it, then sues everybody who does the same thing (or creates product that can talk to this invention). Why? Greed. That's why. Why don't they instead first prove that the technology is their invention, then notify anyone who might be violating their rights, and offer a chance to work out a solution - licence the technology. That way, you get a more permanent revenue stream. So, exactly how many friends does SCO have these days? If they had proven their ownership of the code in the linux kernel first, and then, in a friendly manner, came with the licencing sceme, their business would have been booming now. But noooo. They they want it all and they want it NOW!

Re:exactly!

[supersandra]

Which is, of course, what patents are really about. Very good point.

Copyrights would be more of a "take your house and use it as my own" deal, yes?

I'm always torn on patents; the idea is good but the system is flawed. People can get patents for very stupid or common things. However, if you have a really great and original idea, it also seems like you ought to be able to make money off of it without a bunch of copycats stomping you out of business. At least for an appropriate period of time, at least.

Re:exactly!

[femto]

I happen to know a little about the current situation as I used to work for the inventors of Cisco's technology.

It's interesting to note that when Cisco bought Radiata (the company that developed their OFDM technology), they *didn't* buy Radiata because of their patent! This was told to me by one of the most senior guys in the company.

Radiata's patent covered the baseband digital systems. Cisco bought the company because of the 5GHz radio chip the company had developed.

This radio chip was ahead of anything else available at the time. It was *NOT* patented. The barrier to entry was the high level of R&D and expertise required to reproduce the chip, not a patent.

Whatever the merits of the patent system, OFDM WLAN is not an invention that was a result of the patent system. Rather it was driven by the vision of the inventors, their desire to make great things (and a pile of money) and their desire to stay ahead of the competition.

In this case, the money and rewards followed from being ahead of the competition, not from owning a patent.

Re:exactly!

[kfg]

Copyrights would be more of a "take your house and use it as my own" deal, yes?

Well no, not really. That would be plagiarizing. Taking an idea and claiming it as your own.

. . .if you have a really great and original idea, it also seems like you ought to be able to make money off of it without a bunch of copycats stomping you out of business. At least for an appropriate period of time, at least.

Given that the idea is a thing and not just an idea, like one click shopping, sure. Jefferson pretty much got it right right off the bat. He felt that patents and copyrights were wrong, and he may have been right, but putting him in charge of implimentation was the perfect thing to do, because being the sort of egalitarian thinker he was he was able to draft a nearly perfect compromise between freedom of the people and protection for the individual.

He even forsaw that it was going to be corporations that caused all the trouble, not intellectual property itself.

What also made him apropo for the role is that he wasn't an "outsider", he was a prolific inventor himself.

"Nature intended me for the tranquil pursuits of science, by rendering them my supreme delight." - Thomas Jefferson

He allowed free use of his inventions, considering them having been made for the betterment of mankind.

When Benjamin Franklin was informed that someone in England was making copies of his stoves he declared that since he had invented them to improve the lot of the average human the copier was simply doing his work for him.

The patent on the cotten gin was the last patent that Eli Whitney ever filed, although he went on to many other inventions, declaring that some inventions are too valuable to be owned.

I certainly don't think there's anything wrong with inventors making money, I'm an inventor myself, but we have to be very careful about defining "invention" and "appropriate" period of time.

Even if it "costs" me money. I recognize that I'm not the only person in the universe, nor the only person with rights.

KFG

Re:exactly!

[squiggleslash]

One of the flaws inherent in patenting, as opposed to copyright, is that you can violate a patent even if you didn't copy it. That is, if you independently come up with the same idea, you are still in violation of the pre-existing patent.

I thought that worth mentioning as that's another argument against patents that's frequently forgotten and is, to be, the crux of what makes patents, as opposed to copyrights, unjust, though it's not so relevent for this particular example.

Re:exactly!

[kfg]

One of the flaws inherent in patenting, as opposed to copyright, is that you can violate a patent even if you didn't copy it.

You're not a songwriter, are you? We live in daily mortal terror of that very thing.

And it is the very crux of my "build your own house" example.

KFG

Not exactly.

[CaptainAlbert]

> Who cares that it may have cost millions of
> dollars of risk and investment to devise, refine
> and perfect OFDM and the related technologies

Yeah, right.

The concept of Orthogonal Frequency Division Multiplexing is old. Way old. Like, 1960s old. The mathematics behind it could easily be grasped by anyone who knows what a sine wave is. These people certainly didn't devise it. And they admit it, for example in this white-paper:

www.wi-lan.com/library/whitepaper_wofdm_technica l. pdf

If you look at what they're *actually* claiming to own, this W-OFDM technology is really just a bunch of pre-existing technologies - modulation scheme, channel coding, FFTs, embedded pilot channels - which they've lumped together, given a name and patented. If you look at their block diagrams, you'll see little more than an undergraduate textbook on modern communications systems design would show you.

> we just want them to be free for all of us to use,
> so we definitely should bust their patents.

No... we just want unfettered competition to bring us the benefits of the free market, without being bogged down by people claiming to have "invented" things that aren't actually novel in any way.

Re:Not exactly.

[curator_thew]

``If you look at what they're *actually* claiming to own, this W-OFDM technology is really just a bunch of pre-existing technologies - modulation scheme, channel coding, FFTs, embedded pilot channels - which they've lumped together, given a name and patented. If you look at their block diagrams, you'll see little more than an undergraduate textbook on modern communications systems design would show you.``

Nice try, but you're committing the classic mistake: retrospectively assessing an invention. Many inventions look deceptively simple and obvious in hindsight. In fact, it's often the hallmark of a brilliant invention that it's so simple.

Do you think that it was obvious to combine all of those elements in that particular way? Do you think that it required no undue experimentation to perfect the system and reduce it to a workable technology ?

I mean, using your argument, we could say that the transistor was obvious, because it's just a bunch of pre-existing concepts put together.

``without being bogged down by people claiming to have "invented" things that aren't actually novel in any way.``

And equally, being bogged down by people claiming that the invention was obvious, so they can use it themselves to rip off the hard work and cost of the inventor.

Basically, put your money where your mouth is and put more effort into proving that the invention wasn't inventive and non-obvious given the state of the art in 1993.

Re:Not exactly.

[bit01]

Nice try, but you're committing the classic mistake: retrospectively assessing an invention. Many inventions look deceptively simple and obvious in hindsight. In fact, it's often the hallmark of a brilliant invention that it's so simple.

A pretty sounding but faulty argument that patent supporters like to use. Retrospective assessment, by definition, will have more facts available to make a judgment and therefore will be a better judgement.True innovation is obvious both pre- and post- innovation.

Often, so-called innovation is merely an idea whose time has come that will be invented independently in a short period of time by many people with no so-called "prior art". None. The the patent office gives a monopoly to one "inventor", sometimes giving them a multi-million dollar advantage, and penalises many others who've done exactly the same thing. Yet another example of how unfair the patent system is.

If the patent system truly reflected the reality of IP invention rather than some lawyer fiction at a minimum it would allow multiple near simultaneous invention. It would also not make "prior art" the definition of innovation but "obvious to an expert in the field" (not some patent office non-inventor) instead. In addition it would also assume that simultaneous inventors are innocent of copying until proven guilty with a chain of evidence, like most law.

---

It's wrong that an intellectual property creator should not be rewarded for their work.
It's equally wrong that an IP creator should be rewarded too many times for the one piece of work, for exactly the same reasons.
Reform IP law and stop the M$/RIAA abuse.

Re:Not exactly.

[PsiCTO]

Hi all,

In general, and especially when it comes to software patents and so-called business model patents, I'm on the side of busting 'em up. I've seen far too many far too obvious software patents in my work over the past 10 years.

However, I have to come down on the side of Wi-LAN for this one. Like the previous poster, I too am an EE (Comp. E, and Ph.D. EE). But unlike him I have extensive background in this particular matter because as VP of Engineering for Wi-LAN in the late 90's I did a lot of the hands on work making the technology work, wrote ancillary patent applications, and in so doing went through all the history of the technology. The key, novel claim for W-OFDM is the idea of employing OFDM in a wideband signal. This combats fading because each of the sub-carriers effectively experiences flat fading. Prior to that idea (which the authors of the patents had in 1990 or so), OFDM was used in narrowband (relatively speaking) systems and had it's share of problems.

"Just pick up any book on RF coms. This is basic stuff for anyone experienced in the field. "

People are right in that it's now standard textbook stuff. But before 1999, I don't think it was mentioned in any undergrad textbooks, at least none that I'm aware of. The best reference was Cimini's paper (but I digress...).

Other claims/ideas from Wi-LAN related to OFDM pertained mostly to making the technology practical to implement. In the early 90s, there was really no way to economically build an OFDM-based modem for commercial/consumer applications. I used a simple measure of dollars/bit/s/Hz to illustrate the relative costs of DSSS vs OFDM. About 1996 or 1997, fabrication processes made OFDM cost competitive. Part of what the founders of Wi-LAN did was to anticipate what would be needed for a commercial OFDM system and build a company to do the research required to be ready for the arrival of cost-effective technology to realize the modem.

Anyway, you get the point.

Last, and perhaps most important. I sat on the TGa and TGb subgroups in the IEEE 802.11 standards meetings and helped to define the standard. As anyone who has helped in IEEE standards definition work knows, the task groups are required to get signed letters from the holders of any IP related to the standard being worked on saying (I paraphrase) that the IP holder promises to license their technology in a reasonable fashion, at a fair rate, and without bias to all parties. Thus, this lawsuit is all about testing the veracity of Wi-LAN's claims, not about grabbing something that people haven't already agreed might be due to them. If they are right, they stand to get a very small percentage of the cost of a chip set in licensing fees (in the semi business, this is always much less than 5%; a "fair" rate). If they are wrong, they are out of luck wrt all that initial investment and long hours of work.

There is one other point for people to ponder... A common thing for the big companies like Cisco to do is to buy smaller companies. Radiata, the original company that Wi-LAN sued, was bought by Cisco and, hence, their involvement. Why do big companies buy little ones? To get the IP. If you read an article in a recent MIT Tech Review on why big companies can't do R&D, you'll find the author's argument that the VC business model is to blame. He argues that VC fund small, really talented groups of people to invent new things with the promise that the founders will get rich. It works pretty well, well enough that many of the best people don't join big labs like they used to in the 60s and 70s, but prefer to assume the risk of a start up. Thus, to get the "research" large companies are compelled to buy startups. From the VC perspective, everyone wins...

Amusingly, I remember being at an IEEE 802.11 meeting in San Antonio and eavesdropping on a senior Cisco wireless engineer talking about how Cisco valued companies for acquisition (this is in 1999). He basically said that they paid L, M, or N millions for each engineer, executive, and technician. Times were good back then ;-)

Sue a Standard?

[Sneeka2]

Why is it possible to sue a company which makes use of an official standardised specification like 802.11g? If something is an open standard, everybody is free to use it, unless some royalty conditions are specifically included, right? If any company has a patent on any part of a technology, it is usually a proprietary solution and not an official specification, right? So, wouldn't you need to sue the committee that approved this specific technology as standard, rather than the individuals using the standard?

Re:Sue a Standard?

[CaptainAlbert]

> If something is an open standard, everybody
> is free to use it

Right. Except that not all standards are open. :(

> If any company has a patent on any part of a
> technology, it is usually a proprietary solution
> and not an official specification, right?

Wrong. And yes, I was horrified when I discovered this too. But it's really common for an industry standard to contain patented technologies. For example, many emerging communications standards are employing Turbo codes (which have been mentioned on Slashdot before, with various degrees of cluefulness). Now these were invented quite recently, by some French researchers, and they perform incredibly well. But in order to implement these standards, you (or your supplier, or their supplier) have to pay royalties.

The approval committees never actually infringe any patents, so they aren't a sensible target for litigation. They are, however, a sensible target for loud and persistent complaints about patents-in-standards. Most of them have vested interests in the big companies who implement the standards anyway, so don't hold your breath for a change of heart which might actually encourage competition in those markets. :-\

Re:Sue a Standard?

[Sneeka2]

So what you're saying is, that even if a specification is approved and it does not have a big fat 'COMPANY XYZ OWNZ THIS TECHNOLOGY, YOU WILL NEED TO PAY IF YOU USE THIS STANDARD!' on it, you'd still need to parse all the difference aspects of the specification and compare them with all filed patents at the USPO? Or the EPO? Or any county's patent office for that matter?

OMFG!

Re:Sue a Standard?

[jokkebk]

> But it's really common for an industry standard to
> contain patented technologies.

And it's also common to try and patent everything you can while the standardization process is under way. As far as I recall, many telecom companies and cellphone manufacturers hold a wide portfolio of GSM-related patents.

And because corporations are doing this, there is a large incentive for others to patent things in the standard, too, so they can swap and not end up paying huge royalties for other companies involved in forging the industry standard.

Cisco's Patent

[femto]

I happen to be one of the university team which developed Cisco's OFDM technology.

FYI here is the patent which covers that work. My name is not on it. At the time, it was the concept of a wireless version of Ethernet that was seen to be novel. Others had low speed networks (packet radio). High speed wireless point-to-point links also existed. As far as we knew, noone had yet tried to build something that was a network AND high speed.

Anyway, that was my understanding at the time. As is usual, most parties were playing their cards close to their chest, so there could have been others. The only other one I knew of at the time was the Bereley InfoPad. I'll be as interested as anyone else to see of the patent survives the challenge.

I don't like the current patent mess, but the Cisco patent at least was real in that itwas not speculative. There was a serious R&D effort behind it (as shown by the fact that product was produced).

Re:Cisco's Patent

[femto]

I've just had a look at Wi-LANs patents. It's interesting to note that Fattouche and Zaghloul are both serious researchers, not lawyers, so it is unlikely that their patents were speculative. In my opinion, this is not a case of extortion.

This will be an interesting battle to watch! I think it is unlikely that either patent will be eliminated due to lack of merit. Rather it will be genuine prior art claims that win the day.

BTW. I'm not employed by either party anymore.

Re:Cisco's Patent

[Bazzargh]

"At the time, it was the concept of a wireless version of Ethernet that was seen to be novel."

the funny part of this is, that Bob Metcalfe based the design of wired Ethernet on the wireless Aloha-net. I seem to remember an interview where he said they originally moved to cable because they couldn't afford the radio links that U. Hawaii had used.

(yeah I know this is what you're referring to as packet radio - I just happen to be easily amused)

Ok as a serious argument though, this comment from one of the Ethernet pioneers is interesting:

"David Liddle, now general partner at U.S. Venture Partners, said Xerox charged a one-time license fee of just $1,000. That's in contrast to the huge fees associated with Token Ring.

Xerox's stipulation was that the technology couldn't be changed -- it had to interoperate with all other Ethernet implementations. "Thus we made a playing field in which we could all thrive and compete," Liddle said"

http://www.byteandswitch.com/document.asp?doc_id =3 4327

Its interesting because its today's argument happening 20 years ago - IBM attempting to turn a token-ring into a cash cow (like today's patent shills) turned people away from it as a standard, and Ethernet won - admittedly with a 'RAND' approach, not a patent-free approach.

Canada? Why bother?

[Anonymous Coward]

Interestingly, this case is being brought in Canada, where the defense needs to prove its case.

Population of Canada: 35 million
Population of the United States: 293 million
Population of Europian Union: 380 million

So, assuming that Cisco had to stop selling in Canada and instead sold in just the United States and Europe (ignoring Asia, Australia, etc., entirely), their sales would decrease by less than 5% (35/708). Wouldn't it be reasonable for them to just ignore this lawsuit, and in the meantime continue selling in Canada? If the government eventually forces them to stop, it'd really be no particularly big loss, except to Canada--who would no longer have access to Cisco technology. Which would therefore make the government unlikely to stop Cisco from selling there. Seems like Cisco holds all the cards, here.

I think they need

[Anonymous Coward]

a good wap upside the head.

Progress?

[Hido]

One small step for the patent holder.
Two steps back for man kind.......

So much for foresight.

SCO of Wireless

[Anonymous Coward]

Wi-LAN is the SCO of the Wireless world and they have tried this before. I was part of a large roll out of their equipment several years ago, there stuff isn't very good but their major problem is that 802.11 has taken their old proprietary market away. I remember too clearly how arrogant they were that 802.11 wasn't a threat and that it would "never interoperate across vendors".

What do you do when you can't adapt, why, you sue the people that can adapt and make the best wireless products. SCO of wireless.

Patent reform should also restrict enforcement

[clusterix]

In addition to limiting what can be patented and better analysis before granting, there must be some economic morale requirements for enforcement.

There are entirely too many IP shell companies out there that do nothing but threaten and harass useful companies without providing commercial products based on the patents themselves. They have no plans to exploit their manufacturing monopoly in any honest way. Instead, they should be required in some form to manufacturer real products utilizing their IP or risk losing enforeability in some way. That may require them to cross-license needed IP as well as seriously limit this entire anti-social/economic lawyer business. It could be possible that plaintifs in patent cases must first prove their manufacturing intent to some law/court derived set of requirements before action is started.

Re:Patent reform should also restrict enforcement

[Halo1]

The only reason they are able to do that, is that the big companies like IBM, Sun etc pushed the patent system further and further into the state it is in now... They loved the fact to be able to get really broad and generic patents (often even with little or no investment), but forgot that this could also be used against them. Now they are whining about so-called patent trolls...

Re:Patent reform should also restrict enforcement

[kfg]

While your idea has merit on its face, one of the results would be to force the small, shed in the yard inventor, and yes, they still exist, into the manufacturing business, which might well prove economically impossible, or, force them into being the sort of corporate holder of patents that is causing so much of the trouble.

Beyond repairing the patent granting system the ultimate solution is to eliminate patents on ideas, and to require that one working model can at least be demonstrated.

I don't know why someone didn't think of this system before.

KFG

Re:Patent reform should also restrict enforcement

[Dashing Leech]

... require that one working model can at least be demonstrated.

While that option has some merit, it also causes problems for the "small shed in the yard inventor" you mention. For example, I have a concept for a certain product that I think is very ingenious. The parts for me to build it probably cost tens of thousands of dollars. I cannot afford to build it myself. I cannot get VC investment because they generally invest in businesses, not products. I would have to come up with a full business proposal with marketing analysis an such, which I don't have the time or money to do. Plus I'm not really that interested in starting a company around the product and I certainly wouldn't want it to be controlled by venture capitalists.

I also cannot afford to patent it myself, which is generally $10K-$20K. And I certainly can't afford to defend my patent. (As we all should know, a patent is only worth as much as you can afford to defend it.)

Really my only viable option is to license the idea to a company that would find it a useful product and develop it from the idea. Unfortunately, they also generally want a working prototype. Even if they recognize the design will work and it is useful, I am not protected by a patent so they can legally steal my idea. A non-disclosure agreement could help somewhat, but it is not as powerful as a patent. All they would need to do is make some minor changes or improvements to my design and they can leave me in the dust.

So right now I don't have any really attractive options. But requiring a working model more than doubles the cost of patenting for me. So it isn't a solution and doesn't help the "small guy".

Why bother with canada?

[autopr0n]

If these guys win in Canada, it won't affect the US products, or anywhere else in the world for that matter. Cisco will still be able to sell products in the US even if Wi-Lan wins.

Canada - Game Theory?

[NigritudeUltramarine]

Although I can see how someone might think the parent post was a troll, it does present a somewhat reasonable strategy, from a game theory point of view, for Cisco ... basically a Grim Trigger strategy. Cisco threatens the Canadian government that they'll pull out of their market entirely if they don't cooperate with them. Cisco doesn't have much to lose, but Canada has a LOT to lose.

Not just Wi-Fi!

[Anonymous Coward]

Uhh oh, looking around my house, what doesn't use this standard in some way? OFDM or varitations like COFDM are used in counless things, the DVB-T digital TV standard, 3G phones (W-CDMA), ADSL, DAB digital radio, DRM digital radio, DECT cordless phones, HomePlug adapters, HomePNA and various other bespoke standard I can't recall.

OFDM was actually invented by the US military as a set up from frequency hopping, in the 80's the France Telecom research labs spent a lot of time developing it into COFDM.

Different for civil and criminal.

[dj245]

At first I thought in Canada you were guilty until proven innocent. But if my short Canadian incarceration has taught me nothing, I learned that they have very different procedures for civil and criminal proceedings. I can't come up with the proper google keywords to get a proper legal document, but if a Canadian says they have innocent until proven guilty then who am I to argue.

I was found innocent of weapons smuggling BTW.

Re:Different for civil and criminal.

[iantri]

Assuming you aren't trolling, Canadian legal procedures are much like the States (AFAIK), where in a civil case the person more likely to be right wins, and (as in almost any free country) in a criminal case, you are innocent until proven guilty.

The only time this is slightly different is under Canada's new (and highly controversial) anti-terrorism laws, which allow police to hold people without charges (much like the new American laws -- PATRIOT act?). This act is falling apart at the seams, as the courts have recently decided that another part of it (secret court hearings) is illegal.

Defence must prove their case?

[julesh]

I don't know much about the Canadian legal system, but I know that Canada is a member of the British Commonwealth, and as such its legal system is based on the British one.

This means that, unless they have specifically enacted a change in the laws on burden of proof, the decision in a civil case like this one ought to be based on balance of evidence; that is, whichever side is most likely to be in the right should win. Nobody needs to prove anything.

Anyone with knowledge of Canadian law want to confirm or deny this?

They Should Not Be Allowed To Inforce This

[jmorey]

In the article in CNET there is the following quote:

"Without our OFDM patents, there would be no
802.11a/g," he said. "We didn't enforce these
patents sooner, because we didn't want to slow
down development in the market. But now that
the technologies are firmly established, we
feel we must protect our intellectual
property."

Since they did not start enforcing their patents when they first discovered the "infringement" they should not be allowed to enforce them now.

Re:They Should Not Be Allowed To Inforce This

[yeremein]

Isn't this enough to invoke the submarine patent protection? By admitting that they did not enforce the patent when they knew there were infinging users, maybe their patent is not enforceable.

Is there such a thing as submarine patent protection? There certainly should be, but I was under the impression that patents, unlike trademarks, do not have to be rigorously enforced to be enforceable.

Company Admits Wiaiting for Popularity

[Bruha]

The company claims "Without our OFDM patents, there would be no 802.11a/g," he said. "We didn't enforce these patents sooner, because we didn't want to slow down development in the market. But now that the technologies are firmly established, we feel we must protect our intellectual property."

http://news.com.com/Cisco+the+target+of+wireless +l awsuit/2100-7351_3-5245505.html?tag=nefd.top

I'm confused...

[pclminion]

Aren't we supposed to hope the big, faceless, powerful corporation (in this case Cisco) gets its ass kicked?

Long live patents! I mean, er... uh... What am I supposed to do, again?

Re:Huh?

[.no.gov]

Yes. I live in Canada. This is true. You are GUILTY until proven innocent. The GRC Corporation enforces this with their many stakeholders, including Microsoft and Sun Java Systems as some of the larger ones.

Re:Huh?

[RoTNCoRE]

I should point out it doesn't work out that way with criminal law...this is a civil case.