Commercial Implementation of a Network Optimization Approach Leveraging Quantum Computing Infrastructure
Overview
NTT DOCOMO has achieved a world first*1 by developing and deploying the "TA-List Optimization Algorithm" (the "Technology") on its commercial network. This breakthrough method utilizes a quantum computing platform based on quantum annealing. The Technology is designed to simultaneously optimize network operations by minimizing two critical types of network traffic: location registration signals (sent by user terminals to update their position) and paging signals (sent by the network to trigger incoming calls). Implementation results show a significant performance boost, reducing location registration signals by up to 65.3% and paging signals by 7.0% simultaneously.
Paging Mechanism
To understand the importance of the Technology, it is essential to first explain the basic paging and location management mechanisms used in mobile networks.
To keep terminals connected and always ready to receive incoming calls, the network must track each terminal's real-time location. This system is primarily managed through two spatial concepts: TA (Tracking Area) and TA-List.
A TA is a group of multiple base stations. When the network needs to deliver a call or message to a terminal, it broadcasts a notification to all base stations within the TA where the terminal was last recorded. This notification signal is called a paging signal.
A TA-List is an even larger group consisting of multiple TAs. If a terminal moves to a different TA after its last communication, it will not be able to respond to a paging signal sent only to its previous TA. In such cases, the network retransmits the paging signal to all base stations within the entire TA-List and waits for a response from the terminal.
Method for locating terminals using paging signals
By grouping base stations in high-mobility areas into the same TA and performing paging on a per-TA basis, the network can locate terminals using a significantly lower number of paging signals. This method of optimizing the allocation of TAs within a TA-List is referred to as "TA Optimization."
Overview of TA Optimization
Location Tracking across TA-Lists
When a terminal moves outside its assigned TA-List, it transmits a location registration signal to notify the network of its move. This self-reporting mechanism allows the network to keep track of the terminal's current location.
A primary challenge in traditional network operations is the trade-off between the volume of location registration signals and paging signals. If the TA-List area is set too small, location registration signals occur frequently as terminals cross boundaries more often. Conversely, if the TA-List is set too large, paging signals for incoming calls must be broadcast over a wider range. In either case, this imbalance inevitably leads to an increased processing load on the network infrastructure.
TA-List and location registration processing
Challenges & Optimization
To resolve this challenge, it is necessary to determine the ideal grouping of base stations (TAs and TA-Lists) that minimizes both location registration and paging signals, based on massive statistical data such as terminal mobility patterns and incoming call history.
In 2024, DOCOMO developed its initial "TA Optimization Algorithm"*1 to optimize the arrangement of TAs within a given TA-List. While this algorithm successfully reduced paging signals, it could not optimize location registration signals because it did not modify the TA-List boundaries themselves.
To configure TAs and TA-Lists in a way that minimizes total network load, all possible combinations of base stations and area settings must be exhaustively calculated. However, because the number of combinations grows exponentially, this level of computation was extremely difficult for conventional computers to handle.
To overcome these challenges, DOCOMO developed the "TA-List Optimization Algorithm" leveraging quantum annealing technology to simultaneously optimize both paging and location registration signals. This breakthrough allows for the precise balancing of these two signals, which are naturally in a trade-off relationship. In practice, this means the system can automatically design the ideal TA and TA-List configurations—minimizing paging signals for incoming calls while suppressing location registration signals along high-traffic transit routes. Because the priority for reduction varies depending on area characteristics and terminal mobility patterns, achieving this customized balance is crucial for network efficiency.
In the TA-List optimization process, the challenge of minimizing both signal types was formulated as a combinatorial optimization problem. This was then converted into Quadratic Unconstrained Binary Optimization (QUBO) format—a structure compatible with annealing-type quantum computers—and solved using quantum annealing technology. As a result, the search for an optimal solution that previously required immense computational power can now be completed in as little as five minutes.
Furthermore, by incorporating operational constraints—such as avoiding TA-List boundaries on major transit routes like railways—the algorithm ensures that TA-Lists are formed along these routes. This effectively reduces the volume of simultaneous location registration signals and mitigates the overall load on the network.
Overview of TA-List optimization
Implementation Results
While the nationwide rollout of this Technology is currently underway, its application in specific areas has already yielded remarkable improvements. Most notably, we confirmed a significant reduction in the volume of location registration signals generated when terminals move across TA-List boundaries, with a 65.3% decrease during daily peak hours following implementation.
Simultaneously, the volume of paging signals—sent from base stations to terminals for incoming calls—saw a 7.0% reduction (base station average) during peak hours.
Typically, reducing location registration signals leads to an increase in paging signals due to their trade-off relationship; however, the Technology successfully reduced both simultaneously. By reallocating the radio resources freed up by this reduction in control signals to user data traffic, an improvement in communication speed can be expected.
Reduction effect on location registration signals when implementing this Technology in a commercial network
Reduction effect on paging signals when implementing this Technology in a commercial network
Conclusion
Through this initiative, we have demonstrated that quantum computing can significantly contribute to the enhancement of network operations. Moving forward, we will continue to advance our research and development of quantum computing infrastructure, striving to solve a wide range of social challenges through innovative technology.
- As of February 13, 2026, according to NTT DOCOMO research
Person in Charge
