5G RAN Developments: A Look at Licensing Defenses

Changes in the 5G radio access network (RAN) – and in particular network disaggregation – could make the monetization of patents essential to RAN equipment and operations (5G RAN SEPs) more challenging. Some of the challenges arise from identifying directly liable infringers in a disaggregated, multi-component system. These challenges are the subject of the second part of this three-part article. This final part of the three-part series addresses other challenges to the monetization of 5G RAN SEPs such as the implementation of parts of the 5G RAN using open source software, patent exhaustion, and vendor indemnifications. Read Parts I and II.

5G RAN Disaggregation and Standard Configurations

Under the direction of ETSI’s NFV industry group, the RAN is being disaggregated from base stations and controllers into centralized units (CU), distributed units (DU), and radio units (RU), each performing discrete functions with respect to the 5G baseband. The goal is to allow for dynamic flexibility in allocating and controlling resources by concentrating intensive baseband computing into a small number of CUs and locate, for instance, latency-intensive processing into DUs and beamforming and digital-analog conversion in RUs closer to the antennas. It is anticipated that each 5G network operator will tailor its configuration of CUs, DUs, and RUs to the unique geographic and population distribution characteristics of its market. For this reason, the mix of RAN equipment and vendors will vary from operator to operator.

Although baseband processing can be split into any number of configurations, O-RAN, an alliance of operators and component and equipment vendors, and OpenRAN, a project group designing and building RAN solutions, are designing, specifying, and building RAN configurations based on operator requirements using vendor-neutral hardware and software solutions.

One of the areas successfully specified by O-RAN is a class of RUs built in accordance with the so called 7-2 split of the baseband physical layer processing. The functionality of these RUs is narrowed as shown below to discrete activities. The bulk of baseband processing in such a 7-2 split is performed by a combination of DUs and CUs located far away from the actual antenna sites. In such configurations, no one piece of equipment performs all the functions of the RAN – RUs in particular simply receive data streams from hierarchically superior equipment, perform their assigned tasks to such streams, and send the converted analog streams to antenna arrays.

Licensed Use and Exhaustion

Due to the nature of RAN inventions (eg, communications generally require a receiver and transmitter), many of the 5G RAN standard-essential patents (SEPs) apply to both RAN network equipment and to end user terminals. This commingling of end-user terminal claims, infrastructure equipment claims, system claims, and method claims can give rise to exhaustion and license defenses.

In the past, traditional vendors of RAN equipment categorically did not license their patents to operators, but instead authorized operators’ use of their patents through the unconditional sale of their equipment. Patent exhaustion in such cases was acceptable to vendors that could foresee the connection between their patented device claims and the use scenarios for their aggregated RAN equipment and charge appropriately to capture the value. With the disaggregation of RAN functionality amongst a number of components sourced from different vendors, it becomes more difficult to capture the value of 5G RAN through the pricing of narrow functioning equipment that may only contribute to infringement.

The efforts of some vendors to restrict the effect of exhaustion by conditioning their equipment sales on use restrictions has been frustrated most recently by the U.S. Supreme Court in its 2017 decision in Impression Products vs. Lexmark. Here the court found that a patentee’s sale of hardware embodying a device claim exhausts such claim regardless of any contractual restrictions on the use of the hardware. The court did suggest though that, while the patentee-vendor could not enforce its patent rights, it could seek contractual damages from its customers. As pointed out earlier in the second part of this article series, that may not be palatable for patentee-vendors that hope to sell equipment.

The U.S. Supreme Court broadened the effects of patent exhaustion in 2008 in Quanta vs. LG by applying the doctrine to the sale of components that do not as such infringe all the elements of a  method claim but which have no reasonable non-infringing use. For exhaustion to apply the defendant must show that the component substantially embodies or, in other words, embodies all the inventive aspects of, the method claim, as well as has no reasonable non-infringing use. A scenario could be envisioned where a vendor sells a DU to an operator that implements substantially the inventiveness of a method claim but leaves out an element of the claim that, though inconsequential to the invention, is implemented at high cost in an antenna array and RU combination outside the DU that is purchased by the same operator. Under Quanta vs. LG, the unconditional sale of the DU could exhaust the method claim with respect to the downstream use of that DU.

One of the questions left open by Quanta vs. LG was whether the exhaustion of a method claim would also exhaust other claims in the same patent. The Federal Circuit in its 2015 decision in Helferich Patent Licensing vs. NT Times answered this question by requiring that exhaustion be analyzed on a claim-by-claim basis. Though the court found that exhaustion was not triggered with respect to a claim by the licensing and sale of a product embodying another claim in the same patent, the court did not foreclose the possibility of an exhaustion defense with respect to one claim due to the licensing and sale of a product under another claim in the same patent if the claims contain essentially the same inventive features (such as might be found in a paired device-method claim relationship) and the device claim would have to be practiced for the method claim to be practiced. An additional requirement is that the exhaustion would only apply to the use of the product by a so called “authorized acquirer”, i.e., the acquirer at the first sale of the product or another acquirer downstream from the first sale.

Outside an exhaustion defense, patentee-vendors may in certain circumstances be legally estopped from asserting their patent claims against RAN equipment if they earlier had granted a license to such claims to a competitor for valuable consideration and then later sought to derogate from that license. The Federal Circuit found in 2009 in TransCore vs. Electronic Transaction that such an implied license could extend even to those claims of a licensor that were not expressly part of the original grant provided that a license to such claims is necessary to practice the claims covered by the original license. It is not difficult to imagine such a scenario arising from a patentee-licensor’s practice of licensing piecemeal its claims to complementary functionality performed across a number of components in a disaggregated RAN.

Vendors of O-RAN and OpenRAN compliant RAN equipment may be able to show that all the elements of a patent asserted by a RAN equipment vendor are not infringed because some of those elements are licensed under a broad cross license with yet another RAN equipment vendor. For example, a distributed antenna system (DAS) vendor could claim a license defense with respect to an assertion of infringement against its RU if the asserted patent claim contains an element requiring a step provided by a mobile network operator’s DU sourced from a RAN vendor that has a broad cross license under the asserted patent. The propensity of traditional RAN vendors to cross license each other without carefully crafted field-of-use restrictions and the disaggregation of the 5G RAN only exacerbates the problem.

Open Source and Reduced Teardown Costs

Another license defense could arise from the open source software and contribution obligations of contributing members to O-RAN’s and OpenRAN’s specification work. Both of these organizations attract hundreds of members that develop, manufacture, and sell RAN equipment. For example, with the exception of Huawei, all of the traditional RAN vendors are members of the O-RAN Alliance.

The O-RAN Alliance implemented a IPR Policy dated 12/26/2018 that requires members commit to fair, reasonable, and non-discriminatory (FRAND) with respect to their patents necessarily infringed by their contributions and imposes a 60-day review period for specifications prior to their submission for adoption with a withdrawal mechanism for members refusing to license under FRAND with respect to contributions made by others. Failure to timely withdraw from the commitment triggers a license under the preselected terms.

The O-RAN Alliance conducts open source software work focused on aligning a software reference implementation with the Alliance’s open architecture and specifications. For this work, the Alliance requires that each of its members license royalty-free under the Apache 2.0 license any of their patents covering any contribution they make to this general specification implementation work.

With respect to contributions to the Specification Code Project chartered on 7/10/2019, O-RAN requires pursuant to the O-RAN Software License 1.0 that its members that are 3GPP participants license their standard essential patents (SEPs) covering such contributions under the FRAND terms governing the standards body they belong to. A member’s non-SEPs (i.e., patents that are not essential to a 3GPP standard) that are necessarily infringed by its contribution to the O-RAN Specification Code Project are required to be licensed royalty-free under the O-RAN Software License 1.0. The same royalty-free license obligation applies to contributors to the O-RAN Specification Code Project that are not 3GPP participants, regardless of whether their patents necessarily infringed by their contributions are 3GPP SEPs or non-SEPs. The O-RAN Software License 1.0 also imposes a broad, royalty-free grant back under any SEPs owned by a recipient of code resulting from the Specification Code Project.

The Telecom Infrastructure Project (TIP), the umbrella organization of OpenRAN, implemented an IPR Policy with effect as of 5/27/2016. The IPR Policy provides that TIP members agree to the royalty-bearing and/or royalty-free licensing options established by TIP chartered Project Groups prior to their participation. A 60-day review period is triggered on the submission of a Project Group’s draft specification for adoption allowing participants to withdraw their licensing obligation with respect to the reviewed specification. Failure to timely notify the Project Group results in the imposition of the licensing obligations initially agreed to. TIP members may not however withdraw from their initially accepted licensing obligations with respect to any of their patents necessarily infringed by their contribution to the adopted specification. In assessing any license triggered under the TIP IPR Policy, it should be born in mind that pursuant to the IPR Policy the scope of a TIP Project Group expressly excludes any 3GPP standards referenced in the adopted specification.

Despite the difficulties identified above, RAN licensing could be more accessible due to an enhanced ability to prove infringement. This because O-RAN and OpenRAN compliant software has been largely implemented as open source software. This could make it easier and cheaper for licensors to identify and prove infringement. O-RAN and OpenRAN compliant hardware is also significantly cheaper to purchase compared to, for instance, LTE eNodeBs thereby reducing teardown costs.

Unique Indemnification Environment

While vendors of O-RAN and OpenRAN compliant equipment may be able to avoid secondary liability for infringement as discussed above, liability could be self-imposed through the indemnifications required by operators.

Usually, product indemnifications are used to exclude or at least limit liability for infringement of standard essential patents or for infringement caused by others. However, since network operators have been procuring large volumes of expensive equipment for decades they have been successful in pushing understandably uncapped liability on to vendors for infringement of any patent whether standard essential or not. While large vendors with large network supply contracts and cross-licensable patent portfolios have grudgingly been able to accept these unlimited liability demands of network operators, the problem will roost with small vendors of competitively priced RAN equipment. These vendors will find it difficult to refuse the indemnification demands of operators. This will likely lead them to take on liability that far exceeds their capacity to pay. The result could be operators holding seemingly sound vendor indemnifications rendered worthless due to vendor insolvency.

Another result of mobile operators’ significant leverage against traditional RAN vendors has been the prevalent use of combination indemnifications. In essence these indemnifications give operators control over the allocation of damages for any infringement caused not only by the vendor’s equipment but also by the intended use of the vendor’s equipment in combination with any other equipment sourced from third parties. These combination indemnifications curtail any defenses a RAN equipment vendor may have with respect to an operator and create a situation for quasi-divided infringement vis-a-vis the operator and its vendors. In the end, mobile network operators hold the power to allocate infringement damages and settlements across all the vendors in their RAN network and full cooperation is expected.

The Recipe for Success

5G should prove to be disruptive to traditional RAN licensing, giving rise to a more accessible market with a slate of new vendors. Non-practicing licensing entities should find this market attractive provided they are able to source quality patents focusing on discrete functions performed by single entities. Traditional RAN vendors with licensable portfolios are likely finding the new market disconcerting, leaving them with existential questions: assert their portfolios against RAN and DAS vendors and risk disruption to their RAN equipment sales or compete for market share against nimble DAS vendors. Success in the former will require, again, quality patents focusing on single equipment performing discrete functions. Without this, traditional RAN vendors may have to invite their customers to their patent enforcement actions. Success in the latter only promises good things for RAN and DAS operators and eventually end users.

Image Source: Deposit Photos
Image ID:259048638
Copyright:denisismagilov