Lockheed turns 5G networks into drone sensors

Lockheed Martin has demonstrated a drone-detection system that uses existing commercial 5G infrastructure as part of a distributed sensing network, combining radio-frequency measurements, artificial intelligence, and airspace tracking without requiring changes to deployed cellular radios.

The NetSense Airspace Awareness-as-a-Service system was demonstrated in the Miami area in July with Verizon, NVIDIA, Keysight Technologies, ODC, and Lockheed Martin subsidiary Astris AI. The companies used an existing Verizon 5G spectrum environment to detect, alert on, and maintain track custody of unmanned aircraft during the live event.

NetSense analyses changes in radio-frequency signal measurements rather than relying solely on a dedicated radar installed for the protection task. ODC’s AI-native radio-access-network software, NVIDIA AI Aerial technology, Keysight RF simulation tools, and Lockheed Martin warning and tracking algorithms were combined to turn existing communications infrastructure into another source of airspace-awareness data.

The attraction lies in coverage and deployment speed. Mobile networks already place radio infrastructure across cities and around critical sites, so a sensing layer able to use those assets could extend detection without requiring a separate network of bespoke sensors to be installed everywhere from scratch.

The demonstration does not make 5G infrastructure a direct substitute for conventional counter-UAS equipment. Dedicated radars, electro-optical systems, acoustic sensors, and RF detectors provide different information and perform differently depending on terrain, target type, weather, and spectrum conditions. A telecommunications-based sensing layer is therefore more likely to add another source of detections and tracks to a wider architecture than eliminate purpose-built sensors.

The core engineering problem is extracting a reliable security signal from infrastructure designed primarily for communications. Urban RF conditions are complex, with buildings, vehicles, moving people, network traffic, and numerous emitters altering the environment. Algorithms have to distinguish changes associated with an unmanned aircraft from normal network variation while keeping false alarms at an operationally useful level.

That requirement moves substantial work into software, data processing, modelling, and test. Keysight’s role includes RF simulation, digital-twin, emulation, and engineering validation, allowing the partners to reproduce conditions and evaluate algorithms without depending entirely on live drone flights.

Repeatable validation will become more important as deployment expands because the RF environment around an airport, industrial plant, military installation, or urban utility site will not be identical. A system offered as a service needs evidence that its detection performance can be characterised and tuned for different network layouts rather than relying on one successful demonstration.

NetSense also illustrates the increasing overlap between commercial telecommunications engineering and national-security sensing. The system relies on commercial technology and an open architecture, with Lockheed Martin arguing that additional capabilities can be introduced without replacing the underlying sensing infrastructure.

That reduces some barriers to upgrade, but it increases the importance of interface control and cybersecurity. A distributed sensing service connected to commercial networks has to protect the integrity of detections, tracks, alerts, and customer data while operating across infrastructure managed by several organisations. Authentication, encryption, software provenance, patching, monitoring, and access control therefore become part of the airspace-protection architecture rather than separate IT concerns.

The commercial model is also different from conventional counter-UAS procurement. Lockheed Martin plans to offer NetSense as a subscription interfacing with customers’ existing security operations. Pilot deployments are planned for the second half of 2026 and early 2027, with broader commercial availability targeted for 2027 and early deployments focused on selected customers with urgent requirements.

A service model can reduce some upfront infrastructure costs, but it shifts attention towards recurring availability, network resilience, data ownership, and service-level performance. Operators protecting critical infrastructure will need confidence that the sensing layer remains useful during congestion, maintenance, network outages, or cyber incidents rather than only under controlled demonstration conditions.

The architecture also has to define how detections move into an operational response. Finding a drone is only the first stage; security teams need track confidence, identification where available, rules for escalation, and integration with whatever sensor or effector systems are responsible for further investigation or defeat.

Lockheed Martin’s roadmap anticipates future 6G integrated sensing and communications, although the current programme uses available 5G spectrum rather than waiting for a new cellular generation. That gives NetSense a near-term deployment path while preserving a route into networks where sensing may eventually be designed more explicitly into communications infrastructure.

The July event demonstrated that an existing 5G environment can contribute usable drone-detection data without modifying deployed cellular radios. The next test is operational consistency: whether the approach can maintain useful performance across different sites, network configurations, and threat conditions while integrating reliably with the security systems expected to act on its alerts.