When the Cell Tower Becomes Radar: The 5G Race Against Drones

The coverage problem

The hardest part of countering drones is often not destroying them. It is seeing them early enough, across a sufficiently large area, and maintaining a reliable track until someone can decide what to do.

Conventional counter-drone systems usually protect specific sites. Radar, electro-optical cameras, acoustic arrays and radio-frequency sensors can be highly effective around an airport, military base or energy facility. But extending that coverage across an entire city, border or coastline requires many sensors, power supplies, data links and maintenance teams. The cost grows quickly, while the gaps between protected sites remain.

Mobile networks offer a different starting point. Their towers already cover large parts of the populated world. They have power, backhaul connectivity, timing and computing resources, and they are connected to national infrastructure. If those same networks can also sense objects in the air, the cell tower begins to resemble a radar node in a distributed national network.

How a mobile network can detect a drone

The core idea is known as integrated sensing and communications, or ISAC. A 5G or future 6G network transmits radio waves to carry data. Some of that energy reflects from buildings, vehicles, birds and aircraft. By measuring changes in those reflected signals across several sites, the network can estimate an object's position, speed, altitude and direction.

This is different from detecting a drone by intercepting the radio link between the aircraft and its operator. Radio-frequency detectors can be very useful, but they depend on the target emitting a recognizable signal. An autonomous drone, a pre-programmed aircraft, or a platform using an unfamiliar waveform may be much harder to find that way. ISAC seeks to detect the physical object itself, even if it is silent and not registered on the cellular network.

The word “passive” needs care. The drone does not need to transmit, so detection is passive from the target's perspective. But the cellular network is still radiating energy. This is therefore not purely passive radar in the traditional sense. It is better understood as the reuse of communications transmissions for sensing.

From experiment to operational chain

The concept received an unusually relevant test during NATO's recent FUSE 2026 exercise in Latvia. The event brought together roughly 500 personnel from more than 30 countries, around 100 industry partners and almost 400 drones. The scenarios included drone swarms, cross-border incursions, simulated electronic warfare and passive detection using 5G infrastructure.

FUSE was linked to NATO's Digital Backbone Experimentation program, known as DiBaX. The goal was broader than proving that a cellular antenna could notice a drone. NATO tested how resilient communications, counter-UAS sensors, command-and-control tools and targeting systems could exchange data in a contested environment. In practical terms, detection only becomes militarily useful when a track can be fused with other sensors, displayed to the right command post, assessed, and passed to a suitable effector before the target disappears.

Public reporting indicates that Ericsson participated in DiBaX and that Latvian operator LMT served as an integrator. It does not prove, however, that Ericsson's newly announced commercial ISAC product was the exact system used for the 5G sensing element at FUSE. NATO has also not published the detection ranges, false-alarm rates, response times or performance under electronic attack. The exercise demonstrates a serious operational direction, not a fully transparent performance benchmark.

The commercial race

Ericsson says its new ISAC capability can turn existing mobile infrastructure into a distributed sensing system by combining 5G radios, edge computing, signal processing and artificial intelligence. The company claims it can detect small drones at kilometer-scale distances, including aircraft that are not connected to the network. Commercial pilots are planned for 2027, with broader availability expected in 2028. An AT&T demonstration in the United States showed how network data could be used to detect and track airborne objects without requiring a separate radar at every site.

China is already moving from trials toward selected commercial deployments. In Anyang, China Unicom, ZTE and local aviation partners deployed a 5G-Advanced sensing network to support medical blood-delivery flights. The companies said the system monitored more than 20 drones and recorded no detection failures during acceptance testing. In Dalian, a 5G-Advanced deployment around the airport reportedly combined millimeter-wave sensing and edge AI, with vendors claiming detection accuracy of about 98 percent. These figures come from operators and suppliers, so they should be treated as vendor claims rather than independent evaluations.

The concept is not confined to NATO members or China. Telecom operator e& and Tiami Networks demonstrated ISAC in the United Arab Emirates, while Israel has examined cellular-based drone detection for national security needs. The Israeli approach described publicly is not necessarily the same as reflection-based ISAC: it also includes detecting and locating drones through their use of mobile networks. The categories overlap operationally, but they rely on different signatures and have different blind spots.

AI turns reflections into decisions

A cellular network does not receive a clean radar picture. It sees clutter: reflections from buildings, moving vehicles, cranes, weather, birds and countless changes in the electromagnetic environment. Artificial intelligence is needed to separate likely airborne targets from background noise, correlate weak observations from several sites and maintain a track as the object moves between cells.

AI can also fuse the cellular track with conventional radar, electro-optical cameras, acoustic sensors and electronic-support measures. A command system can then estimate confidence, classify the object, predict its route and recommend which sensor should look next. At a later stage it may also propose an interceptor, jammer or other response based on range, cost, rules of engagement and the risk to civilians. The crucial distinction is between recommending an action and authorizing force. The speed of drone attacks increases pressure to automate, but legal responsibility and command accountability do not disappear because an algorithm produced the recommendation.

More autonomy can also make the sensing network resilient. Edge systems may continue detecting and classifying targets when links to a central cloud are degraded. Local nodes can share compressed tracks rather than raw data, reducing bandwidth and latency. But autonomy creates its own risks: a model trained in one city or frequency band may perform poorly in another, and an adversary may try to create false targets or manipulate the data used for classification.

Why 5G will not replace radar

Cellular sensing is attractive because much of the infrastructure already exists, but it is not universal coverage. Performance depends on tower density, frequency, geometry, synchronization, terrain and the local radio environment. Rural areas, mountains, coastlines and war-damaged cities may have sparse or unreliable networks. Low-frequency signals travel farther but provide less precise sensing; higher frequencies can offer better resolution but are more easily blocked.

The infrastructure is also vulnerable. Jamming, cyberattacks, power loss and physical destruction can disable parts of the network. Military use of civilian telecom systems raises questions about privacy, data ownership and the legal status of dual-use infrastructure. A network that contributes directly to targeting may become more strategically valuable, and therefore more exposed, during a conflict.

The most credible future is therefore layered. Cellular sensing will not eliminate the need for radar, cameras or electronic intelligence. It can fill gaps, extend coverage and provide additional observations at a cost that would be difficult to match with dedicated sensors alone. NATO's FUSE exercise is important for precisely this reason. The potential breakthrough is not a single antenna that doubles as radar. It is the connection of civilian infrastructure, AI, resilient communications, command systems and effectors into one distributed defensive architecture.