Comment

Patent trolls, inventors, and Bluetooth designers

What do these companies have in common: Matsushita and its Panasonic unit, Samsung, and Nokia? Two answers: they are all companies named in a patent lawsuit over Bluetooth. And (say wireless engineers) they are also companies where Broadcom would very much like to sell more Bluetooth technology.

That leaves us with a few mysteries.

The first mystery about the Bluetooth lawsuit between patent owner Washington Research Foundation and the above-named companies - who happen to be customers of Cambridge Silicon Radio - is this: if WRF thinks CSR is breaking its patents, why is it not suing CSR?

And a second mystery: why on earth did Broadcom buy into WSF's licence, if CSR thinks it is "without merit"? Is Broadcom falling behind in the technology development world? Or is CSR behaving unethically

I've spoken to technology companies in the past, who have been approached by WRF over patent issues; and as you might expect, not all of them have spoken in praise of the group.

"It's likely they saw some groups of companies, particularly Japanese-owned technology companies, as something of a soft touch," remarked one inventor who has had a brush or two with WSF. "They'd hit the parent company with a host of patent infringement claims, and some of them would be completely unrelated to anything we could see in our portfolio."

Another developer said that "They appeared to think we'd just roll over and sign patent agreements, rather than face their lawyers in court. We saw it as 'money with menaces' really, rather than a serious attempt to show infringement."

It's not a big surprise that companies suspected of infringing patents held by WSF might make unflattering comments about them, but equally, there are plenty of examples of American IP owners which are, fundamentally, legal firms; and plenty of examples of such corporations taking out lawsuits against other IP owners which were settled out of court, rather than endure the losses inherent in fighting. The latest dispute between RIM and NTP  back in April was a good example.

What is a big surprise, is the sight of Broadcom signing a deal with an IP company. It has apparently surprised not only CSR, but the Bluetooth Special Interest Group (SIG). And (as we show below) surprising the SIG on this matter is quite a trick. It thought it had all bases covered, and spent a lot of time and effort to make sure of it. How could it miss a basic patent, crucial to the Bluetooth technology?

What, exactly, are the technology issues in question? Excellent question! - nobody seems willing to answer it.

"To my best understanding," offered one Bluetooth SIG member, "if you start trying to find which patents are being named, you don't get details: but from conversations, I'd say it seems they are the ones about how you tune the oscillators and clever digital IF stuff."

This is not clever stuff that gets written down. "It's about implementations within certain people's chips. Most of that implementation information is kept very secret by the chip manufacturers."

In this instance, rather than fight, Broadcom took out a licence. The obvious implications are either:- 

But there's a third option: that Broadcom decided that a threat of a lawsuit might frighten CSR customers. (We'll dismiss as ridiculous the idea that Broadcom is a "soft touch" for patent trolls and doesn't protect its own IP...)

Few of the engineers I've spoken to believe that either Broadcom, or CSR, had not done work covering those technologies, by the mid-90s. The consensus is that "prior art" exists which would invalidate Suominen's patent - if published. "It's clever stuff, but the basic IP is old, and only the details would be patentable; and everybody has their own way of doing that," said one Bluetooth designer who once worked with Cambridge Consultants in the 90s.

If this is right, there's an interesting sub-question, because nobody can see how WSF might know anything about how the CSR chip handles this implementation stuff. And indeed, nobody I spoke to believes they do. "They're throwing a lawsuit at CSR, in the hope they'll be a soft touch," was what most said.

But of course, WSF isn't suing CSR. It's picking on people who sell Bluetooth based phones and headsets, who happen to be CSR's customers. Why, exactly?

Local paper, Seattle Times quoted John Reagh, WRF's manager of business development and legal affairs: "Instead of suing CSR, he said the organisation decided to act against the handset makers because the chipset manufacturer may not know which chips are headed to the United States, where the patent is enforceable, but the device-maker would."

The US laws of patent are weird enough that the argument makes "sense" in that strange context. But it's hard to avoid the obvious implication: that WSF does not think the patent would stand up anywhere else. And I have to say that it's an argument few people in the UK believe.

What engineers do believe, is rather startling. I don't have any evidence that their opinions are based on any data of any sort, but I can tell you what they're saying to each other in Cambridge

"It's far more likely that when they approached Broadcom, someone smart inside Broadcom saw an opportunity to scare those big buyers of CSR equipment by claiming a patent advantage," said an engineer. "If I were inside Broadcom right now, trying to get a foot inside the Bluetooth door with those people, the most effective threat I could think of would be: 'Buy from us, or you will have your US sales blocked by the Courts while the lawsuit drags on...' and that could be very, very profitable."

Nobody at the Bluetooth SIG is saying anything like that, and nobody at CSR is saying anything other than "the suit is without merit" and "we will defend it robustly" and so on; but it's interesting to discover what the "wireless grass roots" feeling on this is, however accurate or inaccurate it might be.

Much of this grass roots opinion may be pure prejudice, of course. There's a lot of hostility towards patent vampires, especially after the Blackberry case, and even a genuine technology developer like Qualcomm, which actually spent money on R&D to create many of its patents, is experiencing huge hostility from other electronics companies because of its approach to intellectual property.

The Bluetooth SIG approach was to get all the relevant patents needed, and pull all the people who owned them into a sort of commonwealth. Everybody who provided patents could share everybody else's patents - and it's worked rather well, so far.

This approach to intellectual property isn't open to a simple IP owner like Washington Research. Rights to use Bluetooth IP for no charge are only of value for those who are in a position to exploit them technically. If I need to build a headset, then free IP is worth diamonds. If I need to charge royalties on headset patents, however, the last thing I need is free IP.

Washington doesn't generate IP itself. "Our mission is to capture and enhance the value of intellectual property arising from Washington State research institutions to support research and scholarship," it admits frankly. It does incubation; it funds startups using R&D discoveries made in Washington, and it pours revenues back into Washington academic institutions.

But it's not seen, in the UK at least, in the same light as something like the St John's Innovation Centre in Cambridge.

Here's a brief summary of how British technology startups feel about WSF: "I think of WSF as a patent vampire because they're one of many such institutions in the US - a 'research foundation' which seems to own a lot of intellectual property 'on behalf of' other people - patents which were originally owned by companies that were going bust, or by undergraduates who didn't know how to exploit them."

The speaker is a UK entrepreneur. He claims to have been approached by WSF on many occasions, while working for, or establishing, startup technology companies in Europe and in Japan.

His view is that often, they don't even suspect their targets of breaching patents; they simply see it as a commercially viable operation to suggest a deal. "They have some IP in a related field. It would take either a lot of expensive investigation by highly qualified technical people, or a lawsuit, to establish whether or not the IP is being used by the company they are approaching."

So, he thinks, they simply start the lawsuit as a win-win tactic. If the technology is close enough to make the lawsuit expensive, WSF can afford it, and the startup, probably, can't. If the startup defends its claim, it has to fund the disclosure of the evidence that would otherwise cost WSF a fortune in technical fees.

All this is (mostly) speculation and "what people are saying" rather than hard fact, of course. We still don't actually know what patents are being disputed here, so when the Cambridge Mafia says "It's all stuff that doesn't look to be as old as the work which CSR bases its Bluetooth chips on," that's quite possibly guesswork.

What we do know is that the Bluetooth SIG really went a long way down the route of addressing the IP question, and believed it had either established ownership, or done deals with the owners, of all relevant patents.

To quote Al Sacco at CIO, the legal action comes as a surprise to the Bluetooth SIG. Sacco's blog records: "The more than 6,000 companies that belong to the Bluetooth SIG agree to license their relevant patents to other members without cost," and he quotes Anders Edlund, marketing director for the Bluetooth SIG. "It seems to have worked so far, so this [lawsuit] was kind of a surprise," he said.

The WRF claims that at least one of its patents related to radio frequency receivers is used in the Bluetooth standard. Sacco observes, very neutrally and without trying to draw inferences: "That patent was filed in August 2003 and granted Oct. 3, 2006. The first version of the Bluetooth specification was approved in 1998, and several updates have passed since."

Personally, I find it very easy to draw inferences. I think this sounds very much like another abuse of the appalling American patent system.

Only in America could you get a patent in 2006, based on a 2003 application, for work which has been in widespread use for the best part of a decade - and possibly, longer. I really look forward to hearing Broadcom's response to my queries about what patents it paid for in this case, and what made it feel that it could not show "prior art" in basic wireless technology.

And I have to add to my friends inside Broadcom: I recommend maximum disclosure! If you don't defend yourselves from the suspicion that is widespread in Cambridge - that you only signed the patent deal to get a foot in the door at Panasonic, Nokia and Samsung - then I think you're making a serious error of judgement. My advice: be as open as you can about what you're doing, because unless you are, you can expect suspicion to grow.

Here are the questions I sent to Broadcom corporate relations:

No reply, as yet. No doubt, someone in Corporate PR is working on these tricky questions as we speak. I'm not holding my breath...