A team of researchers at Ericsson has found that hybrid quantum-classical computing techniques could improve radio access network (RAN) planning and optimization as mobile networks become more complex and dynamic.

Writing in a technical blog post, the researchers argued that existing heuristic-based planning tools are struggling to keep pace with ultra-dense 5G deployments, increasingly dynamic traffic patterns, and the growing complexity expected in future 6G networks.

The post focused particularly on tracking area (TA) planning, a network optimization problem involving how mobile devices are grouped and tracked as they move between cells.

Tracking areas define the regions in which idle mobile users are tracked. Smaller TAs can increase signaling overhead because devices must update the network more frequently, while larger TAs can increase paging traffic because messages must be broadcast across wider areas.

Ericsson said balancing those competing requirements becomes increasingly difficult as networks scale.

“Modern networks generate optimization problems that grow too large for classical computing to handle efficiently,” the researchers wrote.

The company said quantum computing could eventually help address these combinatorial optimization problems by evaluating multiple candidate network configurations simultaneously rather than relying solely on sequential classical methods.

Because current quantum systems remain limited by noisy intermediate-scale quantum (NISQ) hardware constraints, the researchers proposed a hybrid workflow in which classical algorithms perform preprocessing and baseline clustering before quantum techniques refine potential configurations.

The workflow combines classical clustering approaches such as spectral clustering and Louvain community detection with quantum optimization methods, including quadratic unconstrained binary optimization (QUBO) formulations and variational quantum algorithms. Ericsson evaluated the approach using operational network data and compared hybrid quantum-classical methods against a conventional heuristic baseline.

According to the blog post, a hybrid Louvain-QUBO approach reduced inter-TA handovers by 36.4%, while maximum paging traffic and maximum cells per TA were reduced by 25% and 13.6%, respectively.

The researchers conducted the work using quantum simulation rather than production quantum hardware, though the workflows were designed to reflect hardware-native quantum approaches.

The post reflects growing interest across the telecom sector in whether quantum computing could eventually support network optimization, resource allocation, and autonomous network management tasks.

Large-scale fault-tolerant quantum systems capable of handling production telecom workloads remain years away, and current work in the area remains largely exploratory and research focused.