Network Slicing Patents: Who Leads, Where the Gaps Are 2026
- Filing activity peaked in 2023 at 12 families and has not returned to that level since, suggesting the core slice-allocation claim space is filling up rather than expanding.
- H04W and H04L dominate the classification mix 42 and 25 of 46 records respectively, while AI-driven approaches under G06N appear in only 6 — a small but distinct cluster.
- United States receives the largest share of filings 22 of 46, nearly triple China's 8, with WIPO, EPO, India and the UK splitting a modest remainder.
What this landscape covers
Network slicing design optimization sits at the intersection of radio access network engineering and resource orchestration: patents in this set describe how operators partition a shared network into isolated, SLA-aware slices and allocate radio, compute and transport resources across them. The dataset draws on 46 patent families filed between 2015 and mid-2026 under IPC classes covering wireless network architecture (H04W28, H04W48) and network management protocols (H04L41/08). Most records claim either static template-based slicing or dynamic resource orchestration, with a smaller group applying reinforcement learning to slice allocation decisions.
Because publication typically lags filing by around 18 months, the 2025 and 2026 figures in any trend line understate real filing activity; treat the most recent one to two years as a floor, not a ceiling.
Filing trends and technology composition
The filing curve and classification split point to a field that grew through the early 2020s and has since plateaued, with claim activity concentrated in core wireless and transmission protocol classes rather than spreading into adjacent domains.
A flat-to-declining curve after a 2023 peak
Filings rose from 2 in 2017 to a peak of 12 in 2023. The 2022 midpoint of 6 sits well below that peak, and activity has not sustained its growth rate since — a pattern more consistent with a maturing, already-claimed core than an accelerating one.
H04W and H04L carry the claim weight
H04W (42 records) and H04L (25 records) account for the bulk of filings, confirming that most inventive activity addresses wireless network architecture and transmission-layer management directly. G06N (6) and G06F (5) mark a smaller AI- and software-oriented cluster, while B66B (2) and H04J (1) are incidental classifications rather than a meaningful sub-field.
Shares are the percentage of the 46 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Network Slicing Design Optimization with Eureka
This page is one run against one query. Ask Eureka your own question about network slicing design optimization and every answer comes back with the patent numbers behind it.
Try EurekaThe most-cited records in this dataset
Distributed radio resource orchestration for network slicing (US20260181541A1)
An example method for distributed radio resource orchestration for network slicing includes detecting, from within a radio access network of a communication service provider network, a need of a user endpoint device to access an improved quality of service, defining, in response to the detecting and in coordination with at least one other processing system in the radio access network, a set of radio resources to support the improved quality of service for the user endpoint device, configuring the set of radio resources as a slice of the communication service provider network, and sending an instruction to the user endpoint device that causes the user endpoint device to connect to the slice.Filed by AT&T Intellectual Property I, L.P., published 2026-06-25.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20220086864A1 | Multi-slice support for MEC-enabled 5g deployments | 55 |
| 2 | CN113163451A | 一种基于深度强化学习的D2D通信网络切片分配方法 | 43 |
| 3 | US20210306938A1 | Method and system of template-based dynamic network slicing | 29 |
| 4 | US11115920B1 | Method and system of template-based dynamic network slicing | 21 |
| 5 | CN115460088A | 一种5G电力多业务切片资源分配与隔离方法 | 13 |
| 6 | CN113692021A | 一种基于亲密度的5G网络切片智能资源分配方法 | 13 |
| 7 | US20230007573A1 | Methods and apparatus for media communication services using network slicing | 9 |
| 8 | US20230403731A1 | Multi-slice support for MEC-enabled 5g deployments | 6 |
| 9 | US12289207B1 | Profit-aware offloading framework towards prediction-assisted mec network slicing | 5 |
| 10 | US20230008892A1 | 5g network slicing and resource orchestration using holochain | 5 |
Citation counts reflect influence within the searched corpus and skew toward older filings; treat them as a signal of prior influence, not current filing priority.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. 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 →Reading the citation and geography signals
The most-cited records skew toward MEC-integrated slicing and template-based dynamic allocation, both filed in the US, while a distinct China-origin cluster applies deep reinforcement learning to slice assignment — a technical route that has not yet accumulated the same citation weight.
MEC-integrated multi-slice support leads on influence
The highest-cited record in this set addresses multi-slice support for MEC-enabled 5G deployments, a route that ties slicing directly to edge compute placement rather than radio resources alone.
US filings nearly triple the next jurisdiction
United States filings account for roughly half of the dataset, with China's 8 and a scattered remainder across WIPO, EPO, India and the UK indicating the core claim contest is playing out primarily in US prosecution.
Reinforcement-learning slicing is a small, separate lane
Deep reinforcement learning approaches to slice allocation, exemplified by CN113163451A, form a narrow but technically distinct cluster from the template-based orchestration methods that dominate H04W and H04L filings.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to network slicing design optimization, with the prior art for and against each one.
Assignee landscape and where activity has stalled
Recent-year momentum data shows every tracked assignee, including major telecom equipment and carrier names, at zero filings in the latest year — consistent with the broader post-2023 plateau rather than any single company's retreat.
No assignee shows active filing momentum
Every assignee tracked for recent-year momentum, from carriers to equipment vendors to a university filer, recorded zero filings in the most recent year, with one showing a -100% year-over-year change.
A distinct China-origin RL cluster
China-origin filings cluster around reinforcement-learning-based slice allocation methods, a technically separate route from the template-based orchestration claims that dominate US filings.
A modest, fragmented field rather than a concentrated one
With only 46 families across the full search window, no single assignee has amassed the filing volume typical of a dominant player; the field reads as fragmented across carriers, vendors and a small number of academic and research filers.
| Assignee | Recent year | YoY |
|---|---|---|
| Verizon Patent and Licensing Inc. | 0 | — |
| Intel Corporation | 0 | — |
| Samsung Electronics Co., Ltd. (Korea) | 0 | — |
| Wipro Limited | 0 | — |
| Nokia Solutions and Networks Oy | 0 | -100% |
| Fuzhou University | 0 | — |
| Cisco Technology, Inc. | 0 | — |
| Tata Consultancy Services Limited | 0 | — |
Where to take this analysis
The plateau in filings and the concentration in US prosecution suggest specific next steps for teams deciding where to invest engineering or IP resources.
Check freedom-to-operate against the citation leaders
Before committing engineering resources to MEC-integrated or template-based slicing, review the claim scope of the highest-cited US records directly, since these carry the most established prior art weight.
Explore prior art in EurekaEvaluate the RL-based allocation route separately
The China-origin reinforcement-learning cluster is small and technically distinct from template-based methods, which may leave narrower but still-open claim space for algorithmic slice-assignment approaches.
Run a white space search in EurekaCommon questions on network slicing design optimization patents
This dataset contains 46 patent families published between 2015 and mid-2026 that match network slicing design optimization claims under IPC classes H04W28, H04L41/08 and H04W48. That is a modest total for a telecom sub-field, reflecting a fairly narrow and technically specific search scope rather than the full breadth of 5G network slicing patents generally. Filing peaked at 12 families in 2023 and has not returned to that level since, so the true current total is likely somewhat higher once recent filings publish.
The dataset includes filings from major carriers, equipment vendors and at least one university, but no single assignee shows an outsized share of the 46 total families. Recent-year momentum figures show every tracked assignee at zero filings in the latest year, which is more likely a publication-lag artifact than evidence that these companies have exited the space. Reviewing the full assignee ranking alongside filing dates gives a more reliable picture than momentum alone.
Filings peaked at 12 in 2023 and the 2022 midpoint sat at 6, with no year since matching the peak. Part of this apparent decline is a publication-lag effect: patents filed in 2025 and 2026 typically take about 18 months to publish, so the most recent years in any trend line are understated. The remainder is consistent with a claim space that filled in quickly during the early 5G standardization push and has since become harder to stake fresh, broad claims in.
Template-based dynamic slicing methods, which dominate the H04W and H04L classifications in this dataset, define pre-configured slice profiles that get instantiated and adjusted according to service requirements. Reinforcement-learning-based methods, a much smaller cluster under G06N, instead train models to make slice allocation decisions dynamically based on network state. The template-based route has accumulated more citations and broader filing volume, while the RL-based route remains a narrower, less-contested technical lane.
Based on the classification mix, thin coverage areas include cross-domain SLA arbitration between slices, transport-layer multi-tenant isolation, and energy-aware slice scaling, none of which appear as dedicated clusters in this 46-family dataset dominated by core H04W/H04L allocation claims. Slice-aware handover orchestration is another area with limited direct coverage relative to the core resource allocation methods. Any of these would need a fresh prior-art search focused specifically on that sub-claim before filing, since absence from this search string does not guarantee absence from the broader patent record.
Research Network Slicing Design Optimization in depth with Eureka
Go past this page: query the whole network slicing design optimization 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.