Skip to content
Home/Quantum Computing News/NTT DOCOMO's Second D-Wave Production App Cuts Mobile Network Signaling
IndustryUse Cases

NTT DOCOMO's Second D-Wave Production App Cuts Mobile Network Signaling

DOCOMO is running a second live production application on D-Wave hybrid quantum annealing, optimizing Tracking Area Lists across 333 base stations serving 93 million subscribers.

FreeQuantumComputing
·· 6 min read

On August 18, 2026, D-Wave and NTT DOCOMO announced a second production application running on D-Wave's hybrid quantum annealing service. The first DOCOMO production app, already live, optimized paging inside individual tracking areas and cut paging load by 15 percent. The new app optimizes Tracking Area Lists across cell boundaries, a harder multi-objective problem.

Mobile networks group base stations into Tracking Areas so phones know where to listen for incoming calls. A Tracking Area List assigns each device a set of areas. Shorter lists reduce paging signals but increase location-registration signals when a device crosses a boundary. Longer lists do the opposite. Network planners face a direct trade-off between two costs.

What the deployment claims

DOCOMO applied D-Wave's hybrid solver to a live benchmark of 333 base stations, three Tracking Area Lists, and nine tracking areas. The solver returned a configuration in about five minutes. After integration, DOCOMO reported a 65.3 percent reduction in daily peak location-registration signaling and a 7.0 percent reduction in paging load. The configuration now feeds into DOCOMO's day-to-day network planning and supports over 93 million mobile subscriptions.

Why this matters for annealing

This is not a pilot. D-Wave and DOCOMO describe the application as production-grade and operational. The problem maps cleanly onto quantum annealing: a combinatorial optimization with thousands of binary variables and conflicting objectives. D-Wave's earlier DOCOMO deployment on paging already showed one use case. This second application expands the scope from a single tracking-area optimization to network-wide signaling.

What to keep skeptical

The 65.3 percent and 7.0 percent figures are supplier-and-customer-reported, not independently reproduced. There is no published classical baseline in the announcement, so readers cannot tell whether a well-tuned classical solver would achieve a similar result. The numbers describe a specific workload on DOCOMO's network, not a general proof that quantum annealing outperforms every classical approach on every telecom problem.

For background, see our D-Wave Q2 2026 earnings review, our quantum annealing tutorial, and our roundup of companies using quantum by industry.