From 5G Drone Detection to 6G Network Sensing: Lessons from AT&T and Ericsson

Integrated Sensing and Communication (ISAC), also referred to as Joint Communication and Sensing (JCAS), has featured regularly on this blog. We have previously looked at the underlying technologies, the work of ETSI and the ATIS Next G Alliance ISAC readiness report. A new whitepaper jointly published by AT&T and Ericsson, titled Integrated Sensing and Communication: Why Physical Detection Is Becoming a Network-Scale Capability, provides a different perspective. It uses a live 5G network demonstration to examine what is needed to turn sensing into something operators could eventually offer as a service.

The starting point is a demonstration conducted in July 2026 outside AT&T Stadium in Arlington, Texas. AT&T and Ericsson used three synchronised tower sites near the stadium to detect, locate and track drones flying in authorised airspace at altitudes of approximately 300 to 400 feet (91 to 122 metres). The setup combined Ericsson Massive MIMO radios with sensing-enabled transmissions, dedicated sensing and fusion servers, and AI-assisted signal processing on a live AT&T 5G network. It could track the drones and generate information such as their location, velocity and elevation. The accompanying video also shows how AI-based classification, together with a confidence score, was presented on a live dashboard.

The important point is that the drones did not need to be connected to the mobile network. Unlike a system that relies on a drone reporting its own position, network sensing works by analysing reflections of transmitted radio signals. Changes in the returned signals can reveal an object's distance, movement and direction. Multiple cooperating sites can observe the same area from different perspectives and combine their measurements. This is known as multistatic sensing. It can help overcome some of the limitations of observing an object from a single location, although performance still depends heavily on radio conditions, site geometry and the target itself.

The July announcement attracted attention because it demonstrated an application often associated with 6G using existing 5G infrastructure. However, there is an important distinction between reusing cellular sites and being able to enable sensing everywhere through software alone. The whitepaper notes that converting a tower into a sensing site may require sensing-capable radio equipment, appropriate spectrum and antenna configurations, synchronisation, and dedicated processing. The exact requirements will vary with the environment and the performance expected. In other words, existing infrastructure offers a valuable starting point, but there can still be real investment and integration work.

This becomes clearer in one of the whitepaper's most useful illustrations, which identifies six connected requirements for network sensing. These are a modernised radio and spectrum foundation; sensing-capable equipment; suitable site geometry and sensing topology; standards and capability evolution; processing and control architecture; and service delivery and governance. None is sufficient by itself. Radio reflections need to be gathered across an appropriate area, combined into useful detections and tracks, and then delivered in time to people or systems authorised to use them.

This is where the story becomes more interesting than simply using a cellular network as a distributed radar. The technology might detect an approaching object, estimate its movement and help classify it, but detection is not the same as determining whether it is a threat. A drone is not automatically hostile, and the sensing network does not intercept or disable it. Those judgements and responses belong to security personnel, authorities and their established procedures. Network sensing should therefore be considered an additional source of situational awareness, working alongside conventional radar, cameras, other sensors and relevant identification systems rather than replacing them all.

The potential uses extend beyond major sports venues. The whitepaper discusses airports, critical infrastructure, industrial facilities and transportation environments where awareness of moving objects could improve safety or operations. An electricity substation, a port, an airport perimeter or a busy logistics site may benefit from sensing over an area that is awkward or expensive to monitor using isolated sensors. Nevertheless, the business case will be very different for each environment. The paper acknowledges that a single location may be better served by dedicated sensing, while network-based sensing becomes attractive where wider-area observations, coordination across sites or reuse of existing mobile assets create additional value.

There is also a question of how much the demonstration proves. AT&T and Ericsson showed that three synchronised sites could support detection, tracking and classification under controlled conditions. The public whitepaper does not, however, provide a comprehensive set of numerical results covering detection probability, false-alarm rates, localisation error, sensing latency, maximum effective range or the impact on ordinary mobile-network performance. The reported 300 to 400 feet describes the drones' flying altitude, not the system's maximum detection range. These missing measurements are not necessarily shortcomings of the demonstration, but they matter when assessing whether a system can meet operational requirements reliably across different sites and conditions.

Standardisation is another important part of the journey. 3GPP Release 19 included initial ISAC service requirements, documented in TS 22.137, while Release 20 is progressing further 6G studies and the investigation of sensing capabilities. Release 21 is expected to introduce the first normative 6G specifications. The AT&T and Ericsson demonstration is therefore evidence that useful sensing capabilities can be developed before a complete 6G ISAC ecosystem exists. It is not evidence that a fully standardised, interoperable network-sensing service is already available across commercial 5G networks.

Security, privacy and governance will be especially important if sensing is offered to public authorities or enterprises. A network able to observe objects that are not connected to it raises questions about who may initiate sensing, what results may be shared, how long information is retained, how false positives are handled and who is responsible for decisions based on the data. This also connects with the Next G Alliance ISAC Data Initiative, launched in June 2026, which is examining how sensing information can be packaged, transported and delivered to end-user systems. A useful radio measurement has limited practical value until it becomes trustworthy information in an operational workflow.

The commercial challenge remains open as well. AT&T and Ericsson position sensing alongside the broader evolution towards programmable, AI-enabled networks with distributed processing and high-capacity transport. Those capabilities may help turn observations from several sites into timely information. But operators will still need to define specific services, measurable performance targets, integration with customer systems, support arrangements and workable charging models. A promising proof of concept does not by itself establish who will pay for the service or how widely it should be deployed.

The development of ISAC is not limited to the United States. In September 2026, Ericsson and Finland's Erillisverkot demonstrated 5G ISAC drone detection, including a live trial at Vuosaari Harbour in Helsinki. More recently, in October, Ericsson, Play and Asseco demonstrated drone detection in Poland. The Polish demonstration was particularly interesting because it combined Ericsson's radio-based ISAC technology with Asseco's acoustic drone detection capabilities, bringing information from different sensing methods into a common platform called PUSZCZYK. The demonstration used industrial spectrum for sensing, with Play's 5G network providing backhaul connectivity. These examples suggest that early ISAC applications, particularly for public safety and critical infrastructure protection, could involve integrating network sensing with other detection technologies rather than relying on radio sensing alone.

The larger lesson from the AT&T and Ericsson whitepaper is that ISAC is no longer just a set of proposed 6G use cases or waveform studies. Early multi-site implementations are already being tested on 5G networks. The next challenge is to make their performance repeatable, integrate them into operational decision-making, and develop the standards, interfaces and governance needed for broader deployment. The real measure of progress will not simply be whether a network can detect a drone, but whether it can deliver useful, timely and trustworthy awareness as a sustainable service.

For background, the shorter July demonstration highlights video is also available, alongside the full AT&T–Ericsson whitepaper.

Related Posts: 

Posted by on

Comments

Post a Comment