5G AI/ML Air Interface Patent Landscape 2026
Samsung Electronics holds a commanding lead in 5G AI/ML air interface patenting, accounting for the largest single-applicant position among the top filers, with the field still expanding on a multi-year basis despite annual volume easing from its 2023 peak. Wireless network management (H04W) and AI model computing (G06N) form the dominant technology core, while a handful of adjacent branches remain sparsely covered relative to the field’s overall scale.
Samsung leads a highly concentrated field with a widening tier gap
Samsung Electronics holds the top position with 1,388 patent families, more than four times the second-ranked LG Electronics at 313 patent families. The top five filers collectively account for 70% of the hundred largest filers’ combined total, signalling a heavily consolidated competitive structure.
A sharp tier gap separates the top two players from the rest of the field. Intel (82 patent families), AT&T Intellectual Property (78), and Apple (66) occupy a distinct second tier, while all remaining ranked applicants fall well below 60 patent families each — creating a long, dispersed tail.
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 1 | Samsung Electronics Co., Ltd. | 1,388 | |
| 2 | LG Electronics Inc. | 313 | |
| 3 | Intel Corporation | 82 | |
| 4 | AT&T INTELLECTUAL PROPERTY I L P | 78 | |
| 5 | Apple Inc. | 66 | |
| 6 | Fractal Networks LLC | 58 | |
| 7 | T-Mobile US Inc. | 54 | |
| 8 | Telefonaktiebolaget LM Ericsson (publ) | 42 | |
| 9 | The Massengill Family Trust | 38 | |
| 10 | Boost Subscriberco LLC | 33 |
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 11 | NVIDIA Corporation | 32 | |
| 12 | David E. Newman | 27 | |
| 13 | DISH Wireless LLC | 24 | |
| 14 | China Mobile Communications Group Co., Ltd. | 24 | |
| 15 | Huawei Technologies Co., Ltd. | 21 | |
| 16 | AT&T MOBILITY II LLC | 17 | |
| 17 | Microsoft Technology Licensing LLC | 16 | |
| 18 | IBM Corporation | 14 | |
| 19 | Sempre Inc. | 13 | |
| 20 | Qualcomm Inc. | 13 |
Samsung’s commanding position, reinforced by a 7.3× recent-vs-prior filing surge, suggests the company is actively deepening its air-interface AI/ML portfolio rather than consolidating past filings. This makes it the standard against which any new entrant or challenger must calibrate both freedom-to-operate assessments and technical differentiation strategies.
Filing counts for the most recent 18–24 months are systematically understated due to patent publication lag; apparent softness in 2024–2026 data should not be read as a slowdown. Longer-window growth, applicant concentration, and technology-route coverage are therefore more reliable signals than the latest-year bar alone.
Multi-year growth at 88% with wireless network management as the dominant IPC branch
The filing trend and technology composition charts together show a field that expanded sharply from 2019 onward and is anchored in wireless-network and AI-model classes, with a long tail of adjacent application domains.
Annual filing trend
Filings were negligible before 2019, then surged through a 2023 peak of 718 patent families before easing to 585 in 2024 and 490 in 2025 — figures that are further compressed by publication lag. The field is still expanding on a multi-year basis, with the recent three-year window up 88% versus the prior three-year window; treat post-2023 counts as a floor, not a ceiling.
↗ Hover for values · click a bar to ask EurekaTechnology composition
H04W (Wireless communication networks) dominates with 2,501 records, followed by G06N (Computing based on AI models) at 1,642 and H04L (Digital information transmission) at 1,377 — confirming that the field is tightly coupled between network-layer and AI-layer invention. H04B (Transmission, general) at 724 records forms a secondary core. Classes beyond the top four fall off sharply, reflecting concentrated technical focus rather than broad horizontal coverage.
↗ Hover for values · click a bar to ask EurekaHighly cited patent families surfaced by the query
Citation-heavy patent families returned by the query. Use this section as citation context, not as a curated list of the most topic-specific patents.
Neural network learning apparatus for deep learnin…
Disclosed is a neural network learning apparatus for deep learning and a method thereof. A neural network learning apparatus for deep learning according to an embodiment of the present disclosure includes an input interface, a memory, and a learning processor for applying a Gradient Descent algorithm to a neural network model, and the learning processor may… (excerpt from the patent abstract)
| # | Patent | Citations |
|---|---|---|
| 1 | Spectral sensing and allocation using deep machine… | 549 |
| 2 | Cellular system | 414 |
| 3 | Spectral sensing and allocation using deep machine… | 196 |
| 4 | Spectrum access optimization including for 5G or o… | 155 |
| 5 | System, Device, and Method of Detecting, Mitigatin… | 150 |
| 6 | Method and apparatus for transceiving signal using… | 146 |
| 7 | Cellular system | 108 |
| 8 | Blockchain cellular system | 107 |
Ranked by total forward citations. Citation counts favour older and broadly cited patent families, and broad or adjacent patents may appear when they match the search scope. Treat this section as citation context, not as a curated list of the most topic-specific patents. Some patent titles may be shown in their original, non-English language where an accurate translation could not be guaranteed.
What the competitive structure means for R&D investment decisions
The combination of a high-growth lifecycle, extreme top-applicant concentration, active corporate co-filing, and US-led jurisdiction coverage shapes the risk and entry calculus for any newcomer or incumbent seeking to expand.
Growth stage — multi-year expansion, annual volume past its 2023 peak
The lifecycle evidence classifies this field as Growth: the recent three-year filing window is 88% above the prior period, confirming sustained expansion on a multi-year basis. Annual volume peaked in 2023 at 718 patent families and has eased since, though publication lag means recent-year totals are understated. The field is not yet mature, but the rapid build-up of prior art since 2019 is already constraining claim space in the core H04W and G06N classes.
Growth · post-2023 peakTop-five filers hold 70% of the hundred largest filers’ combined output
Samsung’s 1,388 patent families dwarfs every other applicant, and the top five collectively hold 70% of the top-100 filers’ total. This level of concentration means that any freedom-to-operate analysis must begin with Samsung’s portfolio. New entrants face both a scale disadvantage and a prior-art wall in the core wireless-network management and AI-model classes, making differentiated technical angles or licensing strategies essential.
High concentrationAT&T entities and Samsung-university pairs are the most active co-filers
The most active co-filing relationship is between AT&T Intellectual Property I L P and AT&T Mobility II LLC, with 14 jointly filed patent families — likely reflecting internal IP structuring across corporate entities. Samsung Electronics co-filed 7 patent families with its Beijing R&D subsidiary and 4 each with Sungkyunkwan University and Yonsei University, indicating a deliberate university-partnership strategy to cover foundational AI/ML methods alongside commercial network implementations.
Corporate + academic co-filingUS leads filings; PCT and EPO ensure broad international coverage
The United States is the primary filing jurisdiction, followed by WIPO (PCT) and Europe (EPO), with China and India also seeing notable activity. The breadth of PCT and EPO filings alongside strong US presence indicates that the leading applicants are pursuing international patent protection, not just domestic coverage. India’s position in the top five jurisdictions is notable and may reflect growing local R&D investment and regulatory interest in 5G deployment.
US-led, global reachGo beyond the landscape: Eureka’s TRIZ Solution agent breaks down an R&D problem and returns patented concept solutions, each with a technical approach and cited patent & literature evidence.
| Applicant | Collaborator | Co-filings |
|---|---|---|
| AT&T知识产权一部有限合伙公司 | AT&T移动第二有限责任公司 | 14 |
| Samsung Electronics Co., Ltd. | Samsung R&D Institute China – Beijing | 7 |
| Samsung Electronics Co., Ltd. | Sungkyunkwan University | 4 |
| Samsung Electronics Co., Ltd. | Yonsei University | 4 |
| Samsung Electronics Co., Ltd. | UI (UNIVERSITY IND FOUNDATION) YONSEI UNIVERSITY | 1 |
| Samsung Electronics Co., Ltd. | Seoul National University R&D Foundation | 1 |
Co-filing pairs, ranked by the number of jointly-filed patent families.
Samsung and Intel gaining momentum; LG and Apple sharply receding
Momentum diverges sharply across the top ranked applicants: Samsung and Intel are accelerating, while LG Electronics and Apple have pulled back significantly in recent filings.
Samsung Electronics
With 1,388 patent families and a recent filing surge of 7.3× versus the prior three-year period, Samsung is not just defending its lead — it is actively widening it. The portfolio is concentrated in H04W 24 (wireless network monitoring and management, 1,165 records), H04W 72 (resource management, 384 records), and H04B 7 (transmission, 335 records), confirming a focus on the operational AI layer of the air interface rather than general-purpose AI models. University co-filings with Sungkyunkwan and Yonsei suggest coverage is also being built in foundational methods.
patent families: 1,388Intel Corporation
Intel holds 82 patent families and has recorded a 4.6× increase in recent-period filings versus the prior three-year window — the strongest momentum among the second-tier players. Its portfolio emphasises H04W 24 (66 records), H04W 72 (26 records), and H04L 5 (25 records), reflecting a focus on radio resource management and multi-carrier transmission optimisation. This positions Intel as the most credible technical challenger in the core air-interface AI/ML space, though its total family count remains a fraction of Samsung’s.
patent families: 82| Applicant | Recent (3 yrs) | Trend |
|---|---|---|
| Samsung Electronics Co., Ltd. | 1,019 | ▲ 7.3× vs prior 3-yr |
| LG Electronics Inc. | 7 | ▼ -98% |
| AT&T知识产权一部有限合伙公司 | 21 | ▼ -59% |
| Intel Corporation | 60 | ▲ 4.6× vs prior 3-yr |
| Apple Inc. | 7 | ▼ -88% |
| T-Mobile US Inc. | 32 | ▲ +52% |
| DISH Wireless LLC | 46 | ▲ new entrant |
Under-served branches adjacent to the dominant wireless-AI core
Several IPC classes appear at low share relative to the corpus size, representing areas where current patenting activity is sparse; some carry plausible technical relevance to 5G AI/ML air-interface applications and may warrant monitoring.
G06V · Image and Video Recognition
G06V holds 167 records and just 2% of the top-100 branch share, making it one of the more sparsely covered branches adjacent to the dominant AI core. As 5G networks increasingly carry high-bandwidth video and support computer-vision applications (e.g., intelligent camera nodes, XR streaming, drone surveillance), AI/ML methods for real-time image and video processing at the air interface become relevant. An entrant with existing computer-vision IP could explore how G06V-class methods intersect with H04W resource scheduling to create differentiated filing opportunities.
Search this in Eureka →G10L · Speech and Audio Analysis/Synthesis
G10L appears in 112 records at roughly 1% share, reflecting sparse coverage of speech and audio AI within the 5G air-interface context. Voice-over-NR (VoNR), intelligent codec selection, and AI-driven noise suppression at the radio layer are technically plausible intersections. The low existing density means claim space may be available, though applicants would need to establish a clear technical nexus between air-interface resource management and audio-processing methods to build defensible positions.
Search this in Eureka →How leading applicants differ by technology route emphasis
Route coverage across the main technology branches in the current evidence set.
| Player | H04W 24 · Wireless communication networks | G06N 3 · Computing based on AI models | G06N 20 · Computing based on AI models | H04W 72 · Wireless communication networks | H04B 7 · Transmission (general) |
|---|---|---|---|---|---|
| Samsung Electronics Co., Ltd. | Strong · 1,165 | Emerging · 135 | Emerging · 231 | Moderate · 384 | Moderate · 335 |
| LG Electronics Inc. | Absent | Strong · 243 | Moderate · 81 | Emerging · 25 | Emerging · 7 |
| AT&T Intellectual Property I L P | Strong · 53 | Moderate · 14 | Strong · 27 | Moderate · 21 | Emerging · 10 |
| Intel Corporation | Strong · 66 | Absent | Absent | Moderate · 26 | Moderate · 18 |
| Apple Inc. | Strong · 66 | Absent | Absent | Moderate · 14 | Absent |
| MASSENGILL R KEMP | Strong · 31 | Absent | Moderate · 7 | Strong · 18 | Moderate · 11 |
Frequently asked questions
The corpus covers 3,338 patent families globally. The United States is the leading filing jurisdiction, followed by WIPO (PCT) and Europe (EPO), with China and India also seeing material activity.
Samsung Electronics is the clear leader with 1,388 patent families — more than four times the second-ranked LG Electronics at 313 patent families. Samsung’s recent filing momentum is also the strongest in the field at 7.3× versus the prior three-year period.
The field is classified as Growth. Recent three-year filings are 88% above the prior three-year window, confirming multi-year expansion. Annual volume peaked in 2023 and has eased since, partly due to publication lag in the most recent data.
H04W (Wireless communication networks) leads with 2,501 records, followed by G06N (Computing based on AI models) at 1,642 and H04L (Digital information transmission) at 1,377. These three classes form the technical core of the landscape.
Yes. The most active co-filing pair is AT&T Intellectual Property I L P and AT&T Mobility II LLC with 14 jointly filed patent families. Samsung Electronics co-filed with its Beijing R&D subsidiary (7 families) and with Sungkyunkwan University and Yonsei University (4 families each), indicating a deliberate academic-partnership strategy.
G06V (Image/video recognition) at 167 records and G10L (Speech and audio analysis/synthesis) at 112 records are among the more sparsely covered adjacent branches. G06Q, G06F, and G06K also appear at low share. These are observations of relative sparsity; technical and commercial viability as entry points requires further applicant-level analysis.
Ready to map your own 5G AI/ML air interface landscape?
Join 18,000+ innovators using PatSnap Eureka to map any technology landscape: search 2B+ patents and papers, surface key assignees, and generate a report like this in minutes.
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.