AT&T and Ericsson have demonstrated a way to use deployed 5G infrastructure as a low-altitude drone detection layer, putting network sensing into a near-term telecom enterprise frame.

RCR Wireless News wrote that three commercial 5G sites in Arlington detected and followed a small drone that was not attached to the mobile network.

The demonstration did not depend on installing radar at the venue.

The companies used existing Massive MIMO radios, mid-band spectrum, additional signal-processing software and AI models to infer the drone's position from reflected 5G signals.

5G Sites Become A Passive Sensing Layer

The system used reflections from a non-connected object rather than location information transmitted by the drone itself.

Ericsson combined signal returns from the three sites so multiple tower viewpoints could support the same track and reduce single-site dependence, the RCR account says.

The target environment is low-altitude airspace, where terrain and clutter can make conventional radar coverage difficult.

Robert Soni of AT&T pointed to cell-site placement and to a North American portfolio of about 75,000 AT&T physical locations, giving the carrier a starting footprint that a new dedicated sensor build would struggle to match.

The RCR account says the trial followed test drones at altitudes of 300 to 400 feet and at distances reaching 6 kilometers.

During the demonstration, the system compared the 5G-derived path with the drone telemetry shown on the operator display, and the two paths were reported as closely aligned.

RCR Wireless News described the technical basis as gNB sensing work being standardized in 3GPP Release 19 under Integrated Sensing and Communication.

The standardization work gives Ericsson and AT&T a 5G test path for network sensing before any future 6G deployment cycle.

Reuse Changes The Carrier Business Case

The business argument is reuse.

A dedicated counter-drone radar can offer stronger sensing performance, but a carrier network already has towers, radios and licensed spectrum in place, lowering the extra investment needed for wide-area awareness.

Dell'Oro analyst Stefan Pongratz told RCR Wireless News that enterprise 5G and adjacent services have not yet materially shifted mobile operator results, and that expectations for ISAC are still restrained.

His risk-reward argument is that the existing macro network gives carriers a different starting point because the major assets are already owned.

AT&T is not presenting the work as a finished mass-market product.

Soni described a staged path using skateboard, bicycle and automobile versions, with early demand over the next three to five years expected from government agencies, first responders and critical infrastructure owners.

The article also says federal agencies interested in low-altitude airspace monitoring are helping fund research and development.

AT&T characterized its own internal spending as fairly modest and concentrated in a small Austin technology group.

Classification Remains The Hard Question

The unresolved part is not whether a radio reflection exists; it is whether software can classify that reflection reliably.

The system uses AI inference over the signal-processing output to decide whether the object is a drone and to assign confidence.

The RCR account says public disclosures still omit false-alarm rates, missed-detection rates, detection probabilities by distance and evidence for operation in poor weather.

The missing measurements include the evidence needed to separate drones from birds, weather clutter and site-geometry errors in real operations.

The system also stops at awareness.

Mitigation actions such as jamming, capture or interdiction would need separately authorized counter-UAS processes, so a carrier service would likely feed alerts into a broader security stack.

That distinction limits liability for the network operator but also means customers must integrate sensing output with response procedures, legal authority and other sensors before it becomes a complete airspace-security workflow.

The material evidence gap is field validation: commercial pricing, classification accuracy, false-alarm behavior and handoff procedures remain unproven outside a staged venue demonstration.

For operators, the strongest source-backed claim is that 5G sensing may create an economical detection layer using existing assets, not that it replaces specialized counter-drone systems.