Network slicing refers to dividing a physical network into multiple logical, isolated virtual networks, each tuned to a use case such as massive IoT, low-latency control, or mobile broadband. In this dataset, that concept concentrates in IPC subclass H04W (wireless communication networks) and H04L (digital transmission), distinguishing it from broader 5G filings that cover physical-layer radio design without a slicing or isolation element. A patent search on network slicing should therefore combine RAN-slicing and core-slicing terms with IoT-specific slice language to avoid pulling in unrelated 5G radio patents.

The assignee list in this space includes large telecom equipment vendors and network infrastructure specialists, with names like Nokia, Huawei, Parallel Wireless, AT&T, VMware and Samsung Electronics appearing among the most active filers. None of these assignees show filings in the most recent year in this dataset, which is more likely a publication-lag artefact than an actual pullback, since publication typically trails filing by around 18 months. Filing counts alone should be read alongside patent family size, since raw document counts can be inflated by continuations and multi-jurisdiction filing of the same underlying invention.

The clearest gaps sit outside the core wireless stack: IoT data processing (G16Y), signalling and alarm systems (G08B), aircraft equipment (B64D) and control/regulating systems (G05B) each have only one to four records in this corpus, versus well over a hundred in the core H04W and H04L subclasses. That imbalance suggests slicing concepts have been proposed for these verticals but not yet built into dense claim families. A first filing tying a specific vertical control loop to a dedicated slice identifier, rather than claiming the slicing mechanism itself, has more room to stand.

This dataset covers 280 published records forming 280 patent families across the 2015-2026 coverage window, drawn from a search combining network-slicing and RAN-slicing terms with IoT-slice and machine-type-communication-slice terms. Filing activity rose from 23 records in 2017 to a peak of 56 in 2020, then flattened toward the middle of the period. Because publication lags filing by roughly 18 months, the final year or two in any such count will always look lower than the true filing rate turns out to be.

Not necessarily. Citation counts inside a searched patent corpus tend to favour older filings simply because they have had more years to accumulate citations, and the most-cited record in this dataset, US20170164349A1, was published in the earliest part of the coverage window. A high citation count is a better signal of foundational influence on later filings than of current commercial relevance or legal enforceability. Checking current legal status, family size and geographic coverage is a more reliable way to judge present-day importance than citation count alone.