Diversifying RAN equipment supply
Cellular technologies are highly standardized in the most open and collaborative way, so that any piece of network equipment will work with any device conforming to the same technical specifications from standard-setting organization 3GPP.
However, lack of standardized interfaces supporting interoperability among different functional parts the Radio Access Network (RAN)—as defined by the Open RAN Alliance as RU, Distributed Unit (DU) and Centralized Unit (CU)—has made it difficult or impossible to mix and match cellular network equipment software and hardware from different vendors. Consequently, operators have had to procure all those functional units from the same vendor for any given geographic part of their networks. The O-RAN Alliance that develops the standardized interfaces, has adopted reference designs for outdoor macrocells and indoor small cells (e.g., low PHY layer functionality per 7-2x split).
Open RAN promises to increase choice, competition among vendors and drive down costs in network equipment supply by enabling operators to integrate functional unit products from different suppliers. This approach is already being pursued by new operators including Rakuten in Japan and Dish in the US. Incumbent operators are also major advocates for Open RAN—including the O-RAN Alliance’s founding members AT&T, China Mobile, Deutsche Telekom, NTT DoCoMo and Orange who enjoy questionable governance privileges over others in the alliance. Some of these are making select Open RAN deployments, typically where integration and performance requirements are modest.
Licensing new entrants and establishing FRAND rates
For Open RAN to develop and succeed—as many desire and expect—in addition to technical issues, various commercial issues also need to be resolved. These include patent licensing for the vast trove of standard-essential technologies employed in 4G LTE and 5G network equipment. Given that cellular devices are the most significantly SEP-licensed of any product category, it seems likely that the emerging ecosystem in Open RAN might also become significant in SEP licensing.
Widespread cellular SEP licensing of mobile phones for cash royalty payments took off with the introduction of CDMA-based technologies including CDMAOne in the 1990s, and with CDMA2000 and WCDMA from around the millennium. Leading SEP owners; Ericsson, Nokia and Qualcomm, license these patents, but have ceased producing handsets. Since the introduction of 4G LTE, many different SEP owners’ disclosures have revealed “rate card” licensing charges for their cellular SEPs in handsets and other devices. The leading cellular SEP licensors generate around $9 billion annually.
Previously, in 2G with GSM, an oligopoly of vertically-integrated companies including Alcatel, Ericsson, Motorola, Nokia and Siemens owned most of the SEPs and also produced the standard-compliant devices. The SEPs were either cross-licensed or never asserted among those companies, and so royalty rates were unclear and net payments, if any, were invisible publicly.
The licensing of cellular RAN equipment—where there are also only a small number of vertically-integrated OEMs including Ericsson, Nokia, Huawei—is still rather like the above, as it used to be in 2G for handsets.
Open RAN’s open interfaces are enabling new entrants and new routes to market in cellular network equipment software and hardware supply. However, network function units—including RUs in particular—also implement numerous 3GPP technologies, many of which are SEP based and some of which are different to those in User Equipment (UE). Unlike the vertically-integrated oligopoly of OEMs who already have ingrained access to the SEP technologies required, many of these newcomers will need to license numerous SEPs and will have nothing much to cross-license.
Critical mass for pooling
While neither the operators nor the new-entrant Open RAN technology suppliers are likely to be enthusiastic about making royalty payments to many SEP owners, Alium claims it “provides a one -stop shop to license essential patents and help establish a FRAND rate.”
Patent pooling entirely within the cellular sector has not fared well so far. Bilateral licensing has prevailed between cellular SEP owners and handset OEMs, with few licensors and SEPs licensed through the 3G WCDMA or 4G LTE patent pools. Avanci’s pooling of cellular SEPs for licensing outside the cellular sector to automotive OEMs seems more successful with many licensors and licensees, and pooling holds promise in IoT generally.
Successful pools tend to balance the interest of licensors and licensees. For example, while MPEG LA’s AVC/H.264 video codec patent pool has around 40 licensors, many of these are also major implementers who were motivated to join the pool because they are more interested in minimizing what they are charged to license others’ SEPs than the smaller amounts they can generate in licensing their own SEPs.
In cellular, several major declared-SEP owners have well-established licensing programs and can do better for themselves by licensing bilaterally—particularly if their SEPs are regarded more valuable than the average of those in the pool. Antitrust authorities also demand that bilateral licensing is not precluded by the existence of patent pools, so that competition is preserved.
Participation and rate setting
Although Alium has already revealed its licensing charges, it unusual for a patent pool to do that before disclosing participating licensors. This makes it impossible to estimate what proportion of all applicable SEPs might be included in the pool. Alium would welcome participation from the major cellular SEP licensors, who include infrastructure OEMs; Ericsson, Huawei and Nokia. It seems that rather than expecting those to join, it will be from among numerous others—many of who have far fewer declared SEPs—that participating licensors will emerge.
Alium follows the fashion of setting rates on a dollars-per-unit (DPU) basis. This is most palatable to OEMs producing costly products such as cars, where most of the costs (e.g., for the chassis and tires,) are unrelated to cellular technology. While Alium intends only to license at the RU level, DPU royalties generate the same amount of revenue, regardless of the price of the RU—even if the charge is instead levied on a chip or another component within the RU. Alium’s charges range from $25 down to $10 per RU, depending on unit volumes sold.
The drawback for licensees with DPU royalties is that charges do not reduce when product prices decline. Mobile phone royalties were almost invariably set only as percentages until the mid-2000s because OEMs wanted it that way with the expectation that average selling prices would fall, as they did for a decade or so until then. It was with the introduction and then predominance of smartphones since then that OEMs have demanded royalty caps that turn charges into DPUs for higher-priced handsets.
According to the ABI Research source cited in Alium’s launch announcement, the implied average prices for outdoor macro and indoor micro Open RAN RUs in 2030 will be $4,427 and $194 respectively. That seems to reflect a plausible expectation that indoor small cell RUs will become ubiquitous in the enterprise—like WiFi access points—with 205.5 million shipments forecasted for that year. Corresponding royalties on a percentage basis for incremental sales will be in the wide and unexplained range of a maximum of 0.56% on macros to a minimum of 5.2% on micros. Royalty charges of $10 on RUs selling for around only a couple of hundred dollars are likely to be resisted by OEMs and operators—particularly if the pool’s share of total declared or independently assessed SEPs turns out not to be that great.
While product prices often reduce dramatically as technology gets cheaper and markets grow, FRAND licensing requirements can make it difficult to adjust the basis and level of charging, which are typically set for the life of standards and patent pool programs.
This article was originally published in RCR Wireless.
