05 April 2012

The Story Behind CSIRO’s Wi-Fi Patent ‘Windfall’

windfall: noun 1. something blown down by the wind, as fruit. 2. An unexpected piece of good fortune. (Source: The Macquarie Dictionary.)

Wi-Fi LogoAs has been widely reported in recent days, Australia’s CSIRO has settled outstanding patent claims against US telecommunications companies including AT&T, Verizon and T-Mobile for around $220 million.  The case against these companies has been ongoing for almost two years, although litigation involving the patent in question, which relates to technology underlying all commonly-deployed wireless local area networking (WLAN) systems and components, has a far longer history than this. 

So far, taking into account previous settlements with numerous device and component manufacturers – including HP, Dell, Intel, Microsoft, Netgear, Nintendo, Belkin, D-Link, 3Com and others – CSIRO has reaped over $420 million in income from its WLAN patent.

CSIRO – which stands for Commonwealth Scientific and Industrial Research Organisation – is Australia's national science agency, which lays claim to being one of the largest and most diverse research agencies in the world.  It started life as the Council for Scientific and Industrial Research (CSIR) in 1926.

Many reports in the mainstream media have used words such as ‘windfall’ to describe the CSIRO’s latest victory.  However, as the dictionary definition above shows, a ‘windfall’ is really an unexpected stroke of good luck, often providing something of value at little or no cost.  As the history described below demonstrates, it is hard to imagine a less apt term to describe the success achieved by the CSIRO in securing a return on its significant investments in research and development in radio frequency (RF) technology.

The latest win – and the ones which preceded it – are a testimony to the hard work and ingenuity of CSIRO’s scientists, and the courage, persistence and dedication of its management and legal teams in holding the line, and staring-down some of the largest, most powerful and wealthy technology companies in the world.

SO WHAT’S ALL THE FUSS ABOUT?

Nowadays, we are all generally familiar with WLANs – particularly those operating according to the series of IEEE 802.11 standards commonly known as Wi-Fi.  Indeed, many of us now have such networks in our homes, based on routers than can be purchased for less than $100.  It is therefore very easy to underestimate the significant technical problems which needed to be overcome in order to achieve this modern miracle!

WLANs usually have a network topology consisting of one or more access points with a wired connection to a local area network, and/or to the Internet.  They feature wireless connections to one or more transceivers (i.e. radio transmitter/receivers), most often now built into remote devices such as laptop computers, smartphones, tablet computers, e-book readers, and so forth.  The remote devices communicate with the network access points by way of radio wave transmissions.
The Multipath Problem
A major technical problem encountered in the early days of WLAN development was that of of multiple, echoed signals traveling from transmitters to receivers.  Echoes are caused by transmitted radio waves bouncing off walls and objects within a room or building.  As a result, there are multiple paths, with differing lengths, between any transmitter and receiver, so that echoed ‘copies’ of signals reach the receiver at different times. RF engineers therefore refer to this as the ‘multiple path propagation’ or ‘multipath’ problem.

When the multipath problem is present, a single signal sent from a transmitter will be received multiple times by the receiver over a short period of time. When that happens, to oversimplify somewhat, the echoes from a first transmission may ‘mask’ subsequent transmissions.  To make matters worse, because of the short timescales involved (considering that radio waves travel at the speed of light) this problem becomes more pronounced when higher-frequency signals are used.  But it is necessary to use high-frequency signals in order to achieve broadband connection speeds.

One way to address the multipath problem is to delay the transmission of subsequent signals sufficiently to avoid the masking effect.  However, the time spent ‘waiting’ is essentially wasted, and reduces the maximum data rate the network can achieve.
The CSIRO Solution
In order to combat the multipath problem without reducing the data transmission rate of the system, CSIRO inventors Dr John O’Sullivan, Dr Terry Percival, Mr Diet Ostry, Mr Graham Daniels and Mr John Deane developed a solution that was ultimately described and claimed in its WLAN patent.  Put very simply, their solution was to transmit different portions of a series of signals containing the data over a number of different frequency channels.

By transmitting data in a particular way on many different frequencies, the system would ensure that none of the signals in the series (or their echoes) would interfere with other signals transmitted on different channels.  And because the bandwidth of each frequency channel would be low, it would be largely unaffected by the multipath problem.  Yet the WLAN system could achieve a high overall transmission rate aggregated over the many individual frequency channels.

The US patent in question is no. 5,487,069, entitled ‘Wireless LAN’.  And the invention described and claimed in this patent is so elegant, so effective, and so fundamental to achieving efficient wireless data transmission that no more-practical solution has since been developed.  It genuinely underlies many of the wireless communications technologies we take for granted today.

PRE-HISTORY (BEFORE 1990)

As the CSIRO itself explains:

The invention came out of CSIRO's pioneering work in radioastronomy. That work involved complex mathematics known as 'fast Fourier transforms' as well as detailed knowledge about radio waves and their behaviour in different environments.

Nowadays, fast Fourier transforms (FFTs) barely qualify as ‘complex mathematics’, and for over two decades they have been a standard topic in any good undergraduate telecommunications or electrical engineering syllabus.  This is in part because the digital computing resources required to implement an FFT became well within everyone’s reach.  But before the advent of cheap personal computing, such things were largely the province of mathematicians and scientists involved in such esoteric activities as probing the distant reaches of the universe with radio telescopes.

THE 1990’s

In the early 1990’s, a group of scientists working in the CSIRO’s radio astronomy division turned their minds to the distinctly more terrestrial question of how to make wireless networks work as fast as wired networks.  This led to the invention which was described in patent applications filed in 1992 and 1993.  The US WLAN patent was subsequently granted on 23 January 1996 (which, fittingly, is as close as would have been possible to Australia Day, given that the USPTO issues patents in batches on a Tuesday).

Initially, the CSIRO sought to commercialise the idea itself, via licensing.  In particular, in 1997 Macquarie University Professor David Skellern and his colleague Neil Weste established a company they called Radiata, and took a nonexclusive licence to the CSIRO
WLAN patent with the intention of developing chips implementing the technology.

In 1999, Cisco Systems and Broadcom each invested in Radiata to the tune of $4 million, representing an 11% stake for each investor and valuing the company at around $36 million.  In September 2000, Radiata demonstrated a chip complying with the recently-finalised IEEE 802.11a Wi-Fi standard at a major international exhibition.  It could handle transmission rates of up to 54 Mb/s – as much as many common WLANs achieve to this day.

In November 2000, Radiata accepted a bid of A$567 million from Cisco to acquire the company, making its founders, and many of its 53 employees, quite wealthy – at least on paper.

THE 2000’s

It needs to be said that CSIRO, Radiata and Cisco did not have things entirely their own way.  In particular, a Canadian company called Wi-LAN obtained a patent for a wireless LAN solution based on similar multi-frequency principles prior to CSIRO.  Wi-LAN licensed this technology to Philips and subsequently initiated legal action against Radiata for patent infringement.  A confidential settlement was subsequently reached with Wi-LAN by Radiata/Cisco Systems.

Subsequently, there was significant corporate activity aimed at forming groups
promoting competing Wi-Fi implementations, including the versions backed by Wi-LAN and
also by Cisco and Broadcom using the CSIRO/Radiata solution.  Microsoft also became involved in these manoeuvres.
First Infringement
By 2002, the first unlicensed products containing CSIRO W-Fi technology had begun to appear on the market.  CSIRO still owned the patent, with Radiata/Cisco as a non-exclusive licensee.  It therefore commenced efforts to persuade other wireless chip
manufacturers to pay royalties for their use of the technology, to no avail.

CSIRO then made the brave, and somewhat uncharacteristic, decision to resort to litigation, presumably on the basis that its only other option was to give up and simply allow all of the infringing manufacturers to do as they pleased for free.  And no doubt that is precisely what many of those manufacturers – including some of the biggest names in the business – expected of this comparatively little-known government-funded research organisation from the land down-under.
Litigation Begins
In February 2005, CSIRO fired the first shot, suing Buffalo Technology for patent infringement in the US District Court in the Eastern District of Texas (which had already developed something of a reputation for being a patentee-friendly ‘rocket-docket’ court).  In May 2005, Intel, Dell, HP, Microsoft and Netgear filed counter suits against CSIRO, which therefore suddenly found itself engaged in battle with the ‘big boys’.

In December 2006, CSIRO filed further patent infringement claims against Toshiba, ASUStek, Fujitsu, Nintendo, D-Link, Belkin, SMC Networks and 3Com.  Then in May 2007 Marvell Technology Group sued CSIRO in an attempt to have its patent overturned.  By this stage there was barely a significant manufacturer of wireless components which was not engaged in the dispute with CSIRO.
Injunction
In what ultimately turned out to be a major turning point in the litigation, the District Court in the Easter District of Texas issued a ruling, in favour of CSIRO, in June 2007.  Somewhat surprisingly, the court issued summary judgement (i.e. judgement as a matter of law, without going to a jury trial on the facts), granting CSIRO an injunction barring Buffalo Technology from selling its infringing products. 

The surprise was due, in part, to the fact that only the previous year the Supreme Court had issued a decision in eBay Inc. v. MercExchange, L.L.C. in which it ruled that a patentee is not automatically entitled to an injunction, even if infringement of the patent is proven, where an award of damages and/or a court-imposed compulsory license would be adequate compensation.  In particular, if the patent owner is a ‘non-practising entity’ (NPE) – sometimes disparagingly known as a ‘patent troll’ – then the harm done can only be a potential loss of royalties, and not a loss of sales of a competing product.  Thus an injunction will often not be appropriate.

The District Court was willing to find, however, that CSIRO situation was distinguishable from the MercExchange case, because CSIRO was a legitimate research institution whose reputation and ongoing operations were dependent upon licensing arrangements which included terms other than the purely financial, in order to ensure its good reputation was maintained.

An appeal of the District Court ruling to the Court of Appeals for the Federal Circuit (CAFC) led to a stay of the injunction.  The CAFC panel partially overturned the District Court decision, deciding that the court should not have issued summary judgement on obviousness because there were disputed facts that needed to be put before a jury.  However, all this did was to send the case back to the District Court.
Face-Off
By April 2009, CSIRO was facing-off against 14 companies in a jury trial.  But the writing was on the wall, and it really was a case of who would blink first.  It was apparent that if any of the 14 defendants broke ranks and agreed to settle and pay royalties, it would be immensely damaging to the remaining companies, particularly with the threat of an injunction still hanging over them.

In the event, HP was the first to cave, and the other 13 companies soon followed suit, leading to a reported total settlement of $205 million.  Not many details of the deals were announced at the time, however, because CSIRO still had a few fish to fry.

In September 2009 CSIRO commenced new lawsuits against laptop makers Sony, Lenovo and Acer.  In November chipmakers Atheros and Broadcom joined the action in yet another attempt to have the patent declared invalid. 

THE 2010’s

In March 2010 CSIRO moved further up the food-chain, suing US carriers AT&T, Verizon and T-Mobile for patent infringement.  A trial was set down for April 2012.

Then in May 2010, Marvell finally agreed to settle its dispute with CSIRO.

Which brings us to the most recent events, with around seven defendants agreeing to settle with CSIRO in March 2012 – before commencement of the trial – for a total amount of $220m.  While not all of the companies involved have been identified, they doubtless include Broadcom, Atheros, AT&T, Verizon and T-Mobile.

Unlike past settlements, this one has been widely publicised by the CSIRO and the Australian Government, presumably because it has brought to a close all related legal action over the patent in the US, leaving no matters outstanding before the courts which might limit what CSIRO can say about the cases.

THE END… ?

The US patent is due to expire around the end of 2013, and no doubt CSIRO will be continuing to receive royalties until the very end.

It has not, however, ruled out further legal action, either in the US or any any other jurisdiction in which an equivalent patent was granted.  Certainly by now CSIRO must feel very confident of the validity of its patent rights.  The US WLAN patent has been subject to over a decade of intense scrutiny.  In addition to the attacks mentioned in the above history, it has also been subject to re-examination requests in 2008, by Intel, and in 2010 (twice) by Broadcom.  And it has withstood all of these assaults.

This really has been a David and Goliath(s) battle and – as in the original story, though perhaps less common in real life – David has emerged victorious.

We can now hope that the rewards of all this time and effort are invested in Australian research and innovation, and that some of the results will be equally successful.

Congratulations are certainly due to all involved at CSIRO!

REFERENCES / FURTHER READING

US Patent no. 5,487,069 (via Google Patents).
IP Australia, Wi-Fi patent earns another $220 million.
Senator Chris Evans, Minister for Tertiary Education, Skills Science and Research, Media Release: Aussie scientists bring home millions in wifi windfall.
PatentlyO.com, Nonpracticing Entity (CSIRO) Gets Injunction.
CSIRO, Wireless LANs.
Dr Mark Matthews and Dr Bob Frater, Creating and Exploiting Intangible Networks: How Radiata was able to improve its odds of success in the risky process of innovating, A Case Study Prepared for the Science and Innovation Mapping study of the Department of Education, Science and Training, November 2003 [PDF, 360 kB].

4 comments:

timchild said...

In this particular case CSIRO is a patent troll. It never submitted any proposal to the IEEE, all they have done is go to court based on a patent which itself is based on decades old technology.

timchild said...

That is fair enough and perhaps troll is too stronger statement. However the issue remains, there is no clear indication that they did anything to further what we use today as Wi-Fi. They just patented some technologies, presented a proof of concept and then are expecting to benefit.

Postulative said...

As you'll find at http://www.theregister.co.uk/2012/04/10/csiro_patent_trolls_wifi/, the CSIRO was involved with the IEEE.  It did invent (but not manufacture) new things, as did many other companies with a stake in the WiFi standard that also have patents.  Finally, it defended its recognised patents both directly and through the courts.

Proof of concept?  Rubbish, do your research.  Decades old technology?  In the same sense that Einstein's theory of relativity was built upon Newton's law of universal gravitation.  Nothing comes from nothing - technology develops based on previous inventions, and in that sense the CSIRO improved on what was available.  It was patentable, and defensible, in the same way as the companies it was seeking royalties from charge their own royalties for FRAND patents.

Mark Summerfield said...

Ha! I think I'll leave that one there. No need to spoil the fun for everyone else...

Post a Comment