UnifiedConsulting

5G RAN Licensing
Feb 25, 2021
Proprietary
Proprietary 1
Agenda
● Network of the Future
● Compromises
● 5G RAN Costs
● RAN Disaggregation
● 5G RAN SEP Landscape
● RAN Licensing Challenges
○ Direct Infringement
○ Indirect Infringement
○ License Defenses
○ Exhaustion
● Mitigating IP Risks
Hosts

John Morgan Craig Thompson

Head of Solutions and Deployments General Manager
Facebook Connectivity Unified Consultants, LLC
Founder Telecom Infra Project (TIP)

Proprietary 2
Network Future - 5G to Coexist with LTE and WiFi
Radio 3G WiFi 4 3G HSPA+ LTE LTE-A WiFi 5 LTE-A Pro 5G WiFi 6 WiFi 6E WiFi 7
Standard

Year 1999 2007 2007 2009 2011 2013 2016 2018 2019 2021 2024

6.9G
DL Speed (Max) 7.2Mbps 1.2Gbps 42Mbps 600Mbps 3Gbits 4Gbits 20Gbits 9.6Gbps 9.6Gbps 46Gbps
bps

Latency (Least) 212ms 172ms 98ms 10ms 4ms 4ms 2ms 1ms 1ms 1ms

1/AP 4/AP 8/AP 8/AP

MIMO n/a n/a 2x2 4x4 8x8 256x256 16x16
4x4 4x4 8x8 8x8

Cell Capacity
130/cell 300/AP 130/cell 300/cell 100K/km 1300/AP 100K/km 1M/km 300-1300/AP 1700/AP
(Est)

Mobility (Max) 120kph n/a 120kph 120kph 350kph n/a 350kph 500kph n/a n/a n/a

Cell Size 30km 50-100m 30km 20km 20km 50-100m 20km 100m-5km 50-100m 50m 50m

● 3G, 4G, 5G, WiFi 6, WiFi 6E, and WiFi 7 networks will have to work together
● Different frequencies: low-band (sub 1 GHz), Mid-band (Sub 6GHz, inclusive of CBRS
(3.55-3.7 GHz)), and mmWave-band (24-52GHz)
○ Each with different propagation characteristics and interference challenges

Proprietary 3
Compromises Require a Dynamic, Hybrid Network
● Different uses present different functional
and architectural demands on the network
○ Consumers - speed and availability
○ Gaming - latency and speed
○ Industry 4.0 - reliability and latency
○ V2X - reliability, latency, mobility
● Higher speed and lower latency requires
cell densification and robust fiber optic front
& backhaul
● Long backhaul stretches latency and
reliability forcing data traffic to Edge Servers
at the cost of QoS and security
● Operators are hard pressed to “do it all” -
interconnected private outdoor and indoor
networks are required
Proprietary 4
RAN Costs Driving Architecture and Procurement
● 5G infrastructure investment expected to be
$880B during 2020-2025 (GSMA 2020)
● 50-70% of spend is directed at RAN
● 5G RAN requires >6x more cells and 2-3x
higher energy costs over LTE RAN
● China plans to deploy 4.9M 5G cells by 2030
at an AVG cell spend of $23-25K - US is
forecasting 800K new 5G macro-cells
○ mmW use requires densification
○ Massive MIMO is power hungry
○ Front and backhaul requires more fiber
optic cabling
○ DAS systems are expensive per sqft
A conservative increase of data traffic of just 25%
projected to increase 5G TCO by 60% over LTE
Proprietary 5
Disaggregating 5G RAN
LTE RANs comprise a monolithic baseband unit
● Requiring housing, cooling, and power requirements
close to the antenna array
● Costing >$150K a unit
● RANs are single vendor environments

Untenable for 5G as higher frequency use requires much

higher number of antennas for same coverage

Solution is to split BBU processing

● Accounting for latency sensitivities
● Standardizing interfaces
● Facilitating multi-vendor COTS and SW
● Moving some processing to the cloud

Operators benefit
● Break up vendor lock in, introducing price
competition and more control over updates
● Centralized BB processing reduces TCO
● Densification at a lower cost
Proprietary 6
 Industry Collaborations

5G Standardization NFV ISG Interface Specs OpenRAN Adoption

Proprietary 7
5G RAN Patent Landscape by the Numbers

5G Self-Declared as of 1/2021
● 127K publications
● 37.1K families
● 87.6K (69%) specific to 5G
RAN
● 56% still pending

5G RAN Family Ownership

● 76.3% held by Top 10
● 41.5% held by Huawei,
Samsung, ZTE, NOK, ERIC
● 3% held by NPEs

Proprietary 8
5G RAN Landscape Growth
Self-declared publications 2
years after standard first frozen
● 14K LTE
● 127K 5G
= 9x increase

LTE self-declared landscape

grew 9% annually 2012-2020

With same growth, 5G landscape

could grow to 300K publications
by 2031

Already during 2020, the number

of 5G and 5G RAN self-declared
publications more than doubled

Proprietary 9
5G RAN Licensing Challenges

Disaggregated BB processing and traditional RAN patent claims trigger fundamental questions
● Direct infringement - Which entity is primarily responsible?
● Divided infringement - Which entity controls and benefits?
● Indirect infringement - Is there prior knowledge and intent and what of non-infringing uses?
● Exhaustion - When does substantial embodiment of a process arise in a product?
● TX/RX claim language - Is there a defense if other equipment is licensed?
● License defenses - What triggers an implied license defense?
● Indemnification - How do operators and vendors mitigate their risks?
Proprietary 10
Direct Infringement

Direct infringement is the cornerstone of any infringement case

● “Performed by or attributable to a single entity” (BMC vs. Paymentech (Fed. Cir. 2007))
● “Article containing each claim element” or “all the steps of a process”
● No knowledge or intent required

Operators and private network owners can be identified as being direct infringers of process
claims and system claims requiring the composition of many networked elements
● NPEs will not have qualms over asserting RAN claims against operators and owners
● BUT RAN equipment vendors will find asserting RAN claims against customers unappealing
Proprietary 11
Divided Infringement

HARQ Loop

Attribution of infringement to a single entity can occur where multiple entities involved
● Directs or controls the actions of another - agency or contract (Akamai vs. Limelight (Fed.
Cir. 2015))
○ conditions participation … or receipt of a benefit upon performance of … a patented
method and establishes the manner or timing of that performance.
● Through a joint enterprise where there exists an express or implied agreement, common
purpose, pecuniary interest in the purpose, and equal right of control
● Identifies steps to be taken and other entity performs those steps per prescribed terms in
order to receive a benefit (Travel Sentry v. Tropp (Fed. Cir. 2017))
Proprietary 12
Indirect Infringement - Inducement & Contributory
Inducement and Contributory Infringement require, eg:
● Direct infringement
Antenna Array ● Prior knowledge or no willful blindness regarding the patent
○ DDel concluded in 2019 that patent non-review policy does not,
without more, constitute willful blindness (VLSI Tech vs. Intel)
○ BUT EDTex is mixed: patent non-review policy was sufficient in
2019 for a claim of willful blindness (Motiva Patents v. Sony)
but was not in 2016 (Nonend Inventions v. Apple)
● Intent to cause infringement
○ What’s the significance of vendor’s
Radio Unit
representation of 3GPP compliance?
● Contributory infringement can be rebutted if the
Cloud
accused product has non-infringing uses
○ Defense may apply to COTS gear
Distributed Unit
○ But not necessarily so to special purpose
software
Proprietary 13
Defenses: Exhaustion, Express & Implied Licenses
Licensed Licensed Licensed

Licensed Licensed

Licensed
Licensed

Licensed

Licensed
Licensed

RAN Disaggregation leads to Infringement Fragmentation

If direct infringement provable (and the accused is amenable), conduct the following analysis:
1. Does the asserted claim cover any contributions under OSS or contribution licenses?

2. Is there more than one HW/SW involved in the accused infringement that embodies the
asserted claims or substantially embodies the innovative aspects of the asserted claims?

3. Does the licensor directly or indirectly supply or license such other HW/SW?

4. Does the accused HW/SW have reasonable non-infringing uses?

5. Do the asserted patent claims cover both the accused HW /SW and this other HW/SW?
Proprietary 14
On Licenses: OSS and Contribution Licenses
Licensed Licensed Licensed

Licensed Licensed

Licensed
Licensed

Licensed

Licensed
Licensed

● O-RAN member’s patents covering its contributions to specification implementations are

licensed royalty-free under Apache 2.0
● O-RAN&3GPP member’s SEPs covering its contributions to the Specification Code Project
are subject to FRAND and non-SEPs covering its contributions are licensed royalty-free
under O-RAN Software License 1.0
● Non-3GPP member’s SEPs and non-SEPs covering its contributions to the Specification
Code Project are licensed royalty-free under the O-RAN Software License 1.0
● Use of O-RAN Specification Code triggers broad grant back license to SEPs and non-SEPs
● TIP member’s IP on its contributions is licensed at a min. royalty-free per TIP 2016 IP Policy
● BUT TIP Project Group scope does not cover referenced SSO standards
● TIP member’s failure to oppose a TIP Project Group Licensing Option with respect to that
Project Group’s specification within 60-day Review Period triggers the Licensing Proprietary
Option 15
Still on Licenses: Express and Implied
Licensed Licensed Licensed

Licensed Licensed

Licensed
Licensed

Licensed

Licensed
Licensed

● RAN vendors traditionally cross license their SEPs amongst themselves

○ Cross-license grant and scope is usually defined as all SEPs necessarily infringed by
3GPP standard compliant infrastructure equipment and software
● Licensor may be legally estopped from asserting a patent against an entity if it earlier granted
the entity a license under the patent, received valuable consideration for such, and then
sought to derogate from the license (Wang Labs vs. Mitsubishi (Fed. Cir. 1997))
○ The implied license is between the licensor and the direct infringer
○ Implied license grant may also cover licensor’s patents and claims that were not
expressly part of the original grant if a license to such patents and claims is necessary
to practice the patent subject to the original grant (TransCore vs. Electronic Transaction
(Fed. Cir. 2009))
Proprietary 16
On Exhaustion: Unpatented Components
Licensed Licensed Licensed

Licensed Licensed

Licensed
Licensed

Licensed

Licensed
Licensed

● All claims in a patent that are indistinguishable from an innovation and infringer perspective
may be exhausted with respect to unpatented HW/SW that (i) substantially embodies the
innovative aspects of the claims and (ii) has no reasonable non-infringing use (Quanta
Computer vs. LG (US 2008) and (Helferich vs. NYTimes (Fed. Cir. 2015))
● Substantial-Embodiment test does not apply if the HW/SW is itself patented under the
asserted claims (Keurig vs. Sturm Foods (Fed. Cir. 2013)
● The sale of HW/SW embodying a claim exhausts such claim regardless of any contractual
restrictions on its use (Impression Products vs. Lexmark (US 2017)) - Contractual remedies
may apply
● BUT exhaustion does not occur for separately licensed claims in a patent that are
distinguishable as separate inventions and able to be practiced separately (Helferich vs.
NYTimes (Fed. Cir. 2015)) Proprietary 17
Mitigating Risks
● Operators and system integrators seek indemnifications for SEPs and non-SEPs

○ Often indemnifications cover infringement that arises from the combination of several
devices - the allocation of responsibility for the infringement is often left to the operator

○ Indemnifications agreed with nascent vendors may be meaningless in effect if the

vendor does not have the financial wherewithal to cover - in such cases, breach of an
indemnification can be less costly

● RAN vendors and system integrators are caught between the operators and the licensors

○ They should sharply define the functionality covered by their equipment and services

○ They should contractually limit their liability to that functionality and cap it if possible

Proprietary 18
Thanks for Your Attention

If any questions, please do not hesitate to reach out to either of us.

● John Morgan - [email protected]

● Craig Thompson - [email protected]

Proprietary 19