RAN Intelligent Controller AI Patents: Who Leads, Gaps 2026
- Filings peaked in 2022 at 82 records and have since flattened, suggesting the core RIC architecture claims are largely staked out rather than still being built.
- H04W wireless-network claims dominate at 202 of 208 records while only 36 records sit in G06N, showing most filings protect the RAN integration layer, not the underlying ML models.
- The two most-cited filings share one title "Radio access network intelligent application manager" (WO2023091664A1 and its US continuation) account for 237 and 118 citations, anchoring the field's prior art.
Filing growth compares 2021 (24 records) with 2024 (30) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 208 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This dataset tracks 208 patent families published between 2015 and mid-2026 that combine RAN Intelligent Controller (RIC) architecture with machine learning functions such as traffic prediction, self-optimizing network control, and anomaly detection. The search string ties title/abstract mentions of RIC, Open RAN controllers or RIC platforms to classification codes covering wireless network management (H04W24/02), AI model computation (G06N20) and wireless communication arrangements (H04W28), so the set is narrow to RIC-specific ML claims rather than general Open RAN filings.
Publication activity is concentrated in a single filing office triad: the United States, the European Patent Office and WIPO's PCT route together account for the large majority of receiving-office activity, with India a distant but notable fourth. Because publication lags filing by roughly eighteen months, the apparent slowdown after 2022 should be read with that lag in mind rather than as a definitive peak.
Let an AI agent run this analysis on your own technology
Pick a task. Every answer cites the patents behind it.
Trend and technology composition
The filing curve and the IPC breakdown together show where claim activity actually sits: in RAN-side control logic rather than in the machine learning models themselves.
Filings rose sharply through 2022, then flattened
Annual filings moved from zero in 2017 to a peak of 82 in 2022, the dataset's midpoint year. Volumes since then have not exceeded that peak, consistent with a technology whose core architectural claims (non-RT and near-RT RIC interfaces, SMO orchestration) were substantially filed during the 3GPP and O-RAN Alliance standardization push of 2020-2022.
H04W wireless networking dominates the classification mix
202 of 208 records carry an H04W code, versus 84 in H04L (digital transmission) and 36 in G06N (AI computing). That gap indicates most applicants are claiming how ML output is applied inside the RAN control loop rather than novel model architectures, which sit more often in dedicated G06N filings elsewhere.
Shares are the percentage of the 208 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on RAN Intelligent Controller AI and Machine Learning with Eureka
This page is one run against one query. Ask Eureka your own question about ran intelligent controller ai and machine learning and every answer comes back with the patent numbers behind it.
Try EurekaThe filings anchoring this space
Functional architecture and interface for non-real-time RAN intelligent controller
A service management and orchestration (SMO) entity enabling a functional split between a non-real-time RAN intelligent controller and an external AI/ML server will, during a data collection phase, collect and process RAN and non-RAN data and, during a data transfer phase, transfer that data to the external server via an interface. During a training model input phase, the SMO entity receives a trained AI/ML model, metadata and training results back via an interface.Filed by Samsung Electronics; published 2021-10-28.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | WO2023091664A1 | Radio access network intelligent application manager | 237 |
| 2 | US20240259879A1 | Radio access network intelligent application manager | 118 |
| 3 | US20220167236A1 | Intelligence and Learning in O-RAN for 5G and 6G Cellular Networks | 99 |
| 4 | US20210337420A1 | Functional architecture and interface for non-real-time ran intelligent controller | 47 |
| 5 | US20210029580A1 | Flexible buffer management for optimizing congestion control using radio access network intelligent controlle… | 46 |
| 6 | US20210266907A1 | Facilitation of radio access network intelligent controller resource preservation framework for 5g or other n… | 29 |
| 7 | US20200314622A1 | Beam as a service for 5g or other next generation network | 21 |
| 8 | US20220286914A1 | RIC sdk | 18 |
| 9 | US20220286915A1 | Distributed ric | 17 |
| 10 | US20240098568A1 | Automated detection and mitigation of intra- and interdomain conflicts in open radio access networks | 16 |
Citation counts are drawn from within this searched corpus and favour earlier-filed records; treat them as a measure of influence on later applicants, not of current commercial relevance.
Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the numbers mean for strategy
Three patterns stand out once filing volume, classification mix and citation concentration are read together.
The architecture land-grab has slowed
Filings rose from zero in 2017 to 82 at the 2022 midpoint and have not surpassed that level since. New entrants are more likely to face occupied claim space on core non-RT/near-RT RIC interface architecture than an open field.
Integration claims outnumber model claims six to one
Only 36 of 208 records fall in G06N (AI computing), against 202 in H04W (wireless networks). Applicants are overwhelmingly claiming how predictions and anomaly scores get applied to RAN parameters, not the learning algorithms that produce them.
Two related filings anchor the prior art
WO2023091664A1 and its US continuation, both titled "Radio access network intelligent application manager," carry 237 and 118 citations respectively — far ahead of the next-ranked records. Any freedom-to-operate review in this space should start there.
Filing partnerships are narrow and carrier-centric
Only 10 co-assignee pairings appear across 208 families, and the strongest links sit within single corporate groups rather than across competitors — an operator and its own subsidiaries, or a carrier and its own IP holding entity.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to ran intelligent controller ai and machine learning, with the prior art for and against each one.
Who is filing, and where momentum is shifting
Recent-year momentum figures show a field where even the more active filers are pulling back, not accelerating.
Rakuten group entities are still filing, but slower
Rakuten Mobile and Rakuten Symphony each show a single filing in the latest year, down 50% and 75% year-on-year respectively. Their strongest co-assignee link in the dataset (11 shared filings) confirms tight internal coordination on RIC claims across the group.
Large incumbents show no new filings in the latest year
Samsung Electronics and Intel both register zero filings in the most recent year captured. Given the 18-month publication lag, this likely reflects reporting delay rather than an actual halt, but it means recent competitive signal from these filers is currently invisible in the public record.
AT&T's filing entities have gone quiet
AT&T Intellectual Property I and AT&T Mobility II, which co-file together (6 shared records), show a full year-on-year drop to zero in the latest year. Combined with the broader post-2022 plateau, this points to a filer that front-loaded its RIC claims during the standardization window.
| Assignee | Recent year | YoY |
|---|---|---|
| Rakuten Mobile, Inc. | 1 | -50% |
| Rakuten Symphony, Inc. | 1 | -75% |
| Samsung Electronics Co., Ltd. | 0 | — |
| Intel Corporation | 0 | — |
| RAKUTEN MOBILE USA LLC | 0 | — |
| AT&T Intellectual Property I, L.P. | 0 | -100% |
| Rakuten Symphony Singapore Pte. Ltd. | 0 | — |
| VMware, Inc. | 0 | — |
Where to take this analysis
The dataset points to specific follow-up work depending on whether the goal is freedom-to-operate, licensing, or identifying open filing space.
Run a claims-level FTO check on the top-cited pair
WO2023091664A1 and US20240259879A1 carry the highest citation counts in the set and share a title, indicating a continuation family worth reading claim-by-claim before filing adjacent architecture claims.
Open claim comparison in EurekaTrack the post-2022 plateau against later filing years
Because publication lags filing, the apparent flattening after 2022 needs re-checking as 2025-2026 records continue to backfill into the public record.
Set up a filing alert in EurekaMap the G06N-only subset separately
Only 36 of 208 records sit in G06N; isolating that subset shows which applicants are claiming the ML model itself rather than its RAN integration.
Build a G06N subset view in EurekaCommon questions on RIC AI patents
A RIC is the Open RAN control-plane component, defined by the O-RAN Alliance, that hosts applications (xApps and rApps) making real-time or near-real-time decisions about radio resources. In patent terms, RIC filings typically claim either the architecture connecting the RIC to the SMO and RAN nodes, or the specific ML-driven functions (traffic prediction, anomaly detection, self-optimization) those applications run. This dataset's 208 families split heavily toward the architecture side, since 202 records classify under H04W wireless-network codes versus 36 under G06N AI-computing codes.
By citation count within this corpus, the two dominant records are WO2023091664A1 and its US counterpart US20240259879A1, both titled "Radio access network intelligent application manager," with 237 and 118 citations respectively. Samsung Electronics also holds a heavily cited record, US20210337420A1, covering the non-real-time RIC interface architecture. High citation counts favour earlier-filed records, so this reflects influence on subsequent applicants rather than a live ranking of commercial strength.
Filing volume rose from zero in 2017 to a peak of 82 in 2022, which lines up with the period of most active 3GPP and O-RAN Alliance standardization work on RIC interfaces. Companies typically file core architecture claims once a standard's interfaces stabilise enough to describe concretely, and 2022 appears to have been that inflection point for this dataset. Filing counts since then have not exceeded that peak, though the most recent one to two years are understated because publication lags filing by roughly eighteen months.
The clearest gap is the imbalance between H04W claims (202 records) and G06N claims (36 records): the underlying ML models driving RIC decisions are comparatively under-claimed relative to their RAN integration. Specific under-claimed branches include xApp/rApp conflict resolution logic, non-RT to near-RT RIC model handoff mechanisms, and RAN anomaly root-cause attribution rather than simple detection. Co-assignee data also shows only 10 joint-filing relationships across 208 families, nearly all within single corporate groups, meaning cross-company collaborative filings remain rare and open.
Momentum data suggests a slowdown among the largest named filers: Samsung Electronics and Intel show zero filings in the most recent year captured, and AT&T's two filing entities show a full year-on-year drop to zero. Rakuten Mobile and Rakuten Symphony remain active but at reduced volume, down 50% and 75% year-on-year respectively. Given the roughly eighteen-month publication lag, some of this apparent slowdown will fill in as later filing years are published, so it should be treated as a provisional read rather than a confirmed retreat.
Research RAN Intelligent Controller AI and Machine Learning in depth with Eureka
Go past this page: query the whole ran intelligent controller ai and machine learning corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.