Mallinson: Uncertain future in LTE patent pool licensing

Patent pools can benefit both licensors and licensees but are no panacea for mobile technology licensing.  Innovation and competition are as much about alternative business models as new technology. Patent pooling is just one of various ways of licensing. In fact, antitrust authorities recognise that such collective licensing arrangements can be abused to fix prices and harm competition. Consumers and the market in general are best served by maximizing competition with a variety of Intellectual Property Rights licensing arrangements including pools, bilaterally negotiated royalty payments and cross-licensing.

Pooling patents has been popular in recent years with significant success in standardized audio and video encoding technologies including AAC, MPEG-2 and MPEG-4. Consensus is that these pools have captured the vast majority of IPR owners and their "essential" patents that are needed to implement the respective standards. Licensors have profited by pool administrators maximizing the number of licensees and licensees have benefited from pricing visibility and the efficiencies of buying from a one-stop-shop. Patent pool administrator Via Licensing estimates licensing revenues of around $2bn per year from technology pools--a small figure in comparison to bilateral arrangements.

GSM oligopoly loosened with more varied WCDMA licensing

Rules for pools and licensing deal policies Standards setting organizations such as ETSI, ATIS and 3GPP and trade groups such as NGMN are prohibited by competition (antitrust) law from being the forums for commercial negotiations or setting royalty rates. Neither patent essentiality nor what are deemed to be (Fair) Reasonable And Nondiscriminatory licensing terms are determined by ETSI, 3GPP, NGMN or any other industry association. Patent owners may voluntarily subject themselves to essentiality assessments through patent pooling agreements, but there is no obligation for patent owners in general to do so. Otherwise, determinations only occur in rare instances, such as by a court of law in patent infringement litigation or antitrust cases. ETSI maintains a database of patents that have been declared essential or possibly essential by their owners. NGMN conducted some LTE royalty rate evaluations using a "Trusted Third Party" that totted up maximum royalty rates declared by various companies that claim to own essential IPR for the standard. Aggregate figures were only ever disclosed among members, but are widely regarded as meaninglessly high because they disregard the reality of extensive netting-off of royalty charges that occurs through cross-licensing in bilateral negotiations. It is not NGMN policy to determine patent value on a proportionate basis by counting patents or capping the aggregate LTE royalty at 10% or any other figure. It was a splinter group of five companies, plus their joint ventures, including Ericsson, Nokia Alcatel-Lucent and NEC who first promoted these concepts in their 2008 framework agreement. This approach, which is most appealing to companies who are significantly licensees, has far from universal appeal. For example, RIM rejects the notion of proportionality on the basis that some patents are much more valuable than others. NGMN issued a Request for Information from would-be LTE patent pool administrators. Whereas antitrust approval is essential, neither NGMN nor any other industry body has the authority to determine who can and cannot form a patent pool. One year later, it seems like NGMN is never going to decide who has won its beauty contest. Such a decision would necessarily be highly commercial and involve the conflicting interests of licensors and licensees on matters such as individual and aggregate royalty rates.

Many would-be and actual technology vendors were unhappy with GSM licensing arrangements. They sought something different for WCDMA. In GSM, most of the essential IPR was owned by small group of companies who also manufactured handsets and network equipment. Whereas cross-licensing among these resulted in low or nonexistent IPR charges for the likes of Nokia, Ericsson and Motorola, outside companies suffered from high licensing charges. As a result, handset manufacturers NEC and Panasonic who had enjoyed significant European market shares in analogue technologies in the 1980s were marginalized with the switch to GSM the 1990s.

Two significant differences in essential IPR licensing occurred with the introduction of WCDMA. After selling its infrastructure and handset businesses in 1999 and 2000 respectively, Qualcomm maintained and grew its R&D programs by retaining its chip business and licensing its IPR in CDMA technologies on a widespread basis. In addition, a WCDMA patent pool was launched in 2004.

WCDMA has flourished in terms of market growth and increasing competition. WCDMA technology's global share of handset sales is rising from around 21 per cent in 2009 to around 28 per cent in 2010, according to WiseHarbor estimates. Handset vendor market shares have always been significantly less concentrated in WCDMA than in GSM, with the top five vendors collectively commanding 70 per cent and 80 per cent for each technology respectively in 2009.

Struggling WCDMA pool

The WCDMA Patent Licensing Program patent pool, administered by UK-based 3G Licensing Ltd, has made lacklustre progress since its 2004 inception. It has only incorporated a small proportion of WCDMA IPR. One independently published assessment is that only 10% is included.  This figure is very plausible given that the licensors are predominantly only Japanese companies including NTT, NTT DoCoMo, Fujitsu, Sharp, Mitusbishi, NEC, Panasonic and Toshiba. The rest, apart from Siemens, are operators including France Telecom, KPN and SK Telecom. The notable absence of Qualcomm, Nokia, Ericsson and many other big names makes it seem unlikely anybody else's credible patent count or other evaluation would come up with a much larger percentage.  Mobile is much more complex than audio and video standards with several times more patents and patent owners. It is no surprise that building the pool has been a slog, but problems are fundamental.

Were pool licensors duped by other patent owners who stated intentions also to join, but did not follow through? Whereas Qualcomm has clearly expressed its unwillingness, there were widespread unmet expectations that other major names would join eventually.

What about 4G?

Why should results be any different with LTE? Whereas there is a lot of interest and jostling for position to be the patent pool facilitator and administrator for LTE, there is no reason yet to believe an LTE pool will do any better than the WCDMA pool. The Open Patent Alliance's WiMAX patent pool initiative, administered by Via Licensing,  has similar shortcomings with none of the leading mobile technology patent owners in 2G and 3G--who also lead in IPR ownership for the OFDMA technologies upon which both LTE and WiMAX are based --onboard yet. According to industry analyst and WiMAX proponent MARAVEDIS, there is 60-80 per cent overlap of LTE and WiMAX patents.

Major players in particular continue to believe they can derive more monetary, cross-licensing and litigation defence value by not pooling their patents.  Recent patent disputes, such as RIM with Motorola and Nokia with Apple discourage collective agreements in favour of bilateral defence, despite mechanisms in pooling agreements that enable patents to be asserted in the face of litigation. Significant concessions by existing WCDMA pool licensors would be required to secure a star IPR owner or two. 

While it is unsurprising the three contenders to facilitate and administer one or more LTE patent pools are companies that have already succeeded with other patent pools, it is significant that the current WCDMA pool administrator is not a contender.  MPEG LA, Via Licensing and Sisvel responded to a 2009 Request for Information on forming an LTE patent pool by the Next Generation Mobile Network Alliance.  See sidebar for some basic facts and rules for trade organizations in licensing negotiations.

Sisvel's pitch has popular appeal with regulators, licensees, carriers and consumers who end up paying royalties when they subsidize or purchase handsets. Sisvel founder Roberto Dini made a public conference presentation in London on patents and standards in February 2010 that seeks to show pools can prevent excessive costs from royalty stacking. He suggested that licensed individually, 85 patents for MPEG video at 50 cents apiece would cost $42.50, as opposed to the $2.50 charged by the patent pool. He used several other examples, including the clearly fallacious hypothetical of 954 WCDMA patents at 50 cents each costing $477 in aggregate. Even though around 90% of WCDMA patents are licensed for payment or are cross-licensed bilaterally without pooling, average wholesale selling prices for entire WCDMA phones are only around half that figure.

There is also supposedly a lot of popular support for a WiMAX patent pool. In fact, the pooling of patents in WiMAX is purported to be one of the reasons why royalty costs are touted to be lower with WiMAX than with WCDMA or LTE. A low royalty rate regime could be attractive for Intel with its economies of scale in selling chips, but this makes for poor business by licensors and the administrator. With a weak outlook in WiMAX versus LTE, it is unsurprising that the OPA is now seeking in recent public comments to reposition its initiatives to 4G patent pooling in general, including both LTE and WiMAX.

The OPA has also resorted to scaremongering with demands that Qualcomm should pool its patents to prevent costly legal actions. OPA's President Yung Hahn says that otherwise litigation costs might ultimately be passed along to the end user, thus relegating 4G to the elite, often business users, not the mass market as intended. This is patent nonsense and a repeat of how Nokia, Ericsson, Broadcom, TI, NEC and Panasonic criticised Qualcomm's WCDMA licensing in 2005. It was these companies, not Qualcomm that initiated the associated legal action with the European Commission's competition directorate. As indicated above, WCDMA has since flourished on a widespread basis in a mass market. The complaint failed following years of thorough investigations.

Conflicting interests

Pools for 3G and 4G also face external challenges. Competition authorities are concerned that the pools may be driven by cartel-like activity. Pool rules prohibit the inclusion of non-essential IPR on antitrust grounds, but fathoming out which patents are and are not essential is not straightforward. Most significantly, a pool that is hijacked by licensee and device customer interests might fix low prices to the detriment of licensors and their ability to fund ongoing developments. Some of the minor patent owners in pools and others with strong voices and representation in organizations such as NGMN are more interested in aggregate royalty rate minimization as licensees or device customers than as revenue-earning licensors for the few patents they own, if any. This kind of pricing pressure is also unappealing for patent pool administrators who typically take a percentage of licensing revenues.  

No wonder 3G Licensing Ltd is unwilling to provide the tea and sandwiches as hosts for yet another talking shop on mobile licensing with LTE.