Introduction
Rakuten Mobile has chosen three heavy hitters to help run its next generation mobile core in Japan. Cisco, Nokia, and F5 will support Rakuten’s move to a fully cloud-native 5G Standalone network. The goal is simple to say and hard to do: simplify operations with automation and AI, lower costs at scale, and deliver better experiences for both consumers and businesses. This announcement signals a shift from building a disruptive network to operating it like a high-performance software platform. It is a bet on cloud practices, multi-vendor collaboration, and continuous automation.
What Changed
For the past few years, Rakuten Mobile has been best known for its open, software-centric approach to building a mobile network. The company’s next phase is about operational maturity. Moving to 5G Standalone puts the entire control plane and user plane on a modern service-based architecture, which is how 5G was meant to run from the start. In practical terms, Rakuten will rely on cloud-native core functions, containerized software, service meshes, and automated pipelines that let engineers change, test, and deploy the network like an internet-scale application. The new partnerships formalize who does what. Cisco, Nokia, and F5 will supply key software, networking, and traffic management components that sit at the heart of the 5G core. The selection comes with sharp focus on AI-assisted operations and a cloud-native toolchain from day one.
Why It Matters
A 5G Standalone core unlocks features that Non-Standalone deployments cannot fully deliver. This includes network slicing for enterprises, consistent low latency for real-time applications, and Voice over New Radio for cleaner call quality without falling back to legacy networks. It also sets the stage for automation. In a Standalone world, network functions are microservices that can be scaled, updated, or restarted independently. Fault isolation is cleaner. Observability data is richer. AI models have more meaningful signals to analyze. That operational shift is as important as any speed boost a user might see on a speed test. From a business lens, Rakuten is trying to push cost curves down while increasing the pace of feature delivery. Cloud-native design helps, but only when the integration is done carefully and the production culture matches the technology.
Who The Partners Are and What They Bring
Cisco
Cisco is a familiar name in telco cores and IP transport. It brings mobile core software, segment-routed transport, and a deep ecosystem of automation and telemetry. For a Standalone build, the most valuable pieces are often invisible to end users. Think session management at scale, resilient user plane performance, and policy control that ties subscriber intent to how packets actually move. In a Standalone design, Cisco’s strengths typically show up in how control plane services scale under pressure, how the user plane manages high throughput with fine-grained policy, and how transport features like SRv6 simplify the path between radio sites, core data centers, and the internet.
Nokia
Nokia’s core software portfolio is proven across many operators, and it includes containerized 5G core functions that run on Kubernetes. Nokia’s cloud packet core and data management components are built for dynamic scaling, which fits Rakuten’s goal of running the network like a software platform. Even if the exact workshare evolves, Nokia’s presence gives Rakuten a supplier with deep operational field experience across Standalone deployments worldwide. The timing also aligns with efforts to extend 5G coverage, because a broader radio footprint benefits from a more capable core.
F5
F5 has long served mobile operators with carrier-grade traffic management and security. In a 5G core, F5’s role often appears in the data path where application delivery, CGNAT, DDoS protection, and signaling security live. This is where packet storms are tamed, APIs are defended, and sessions are balanced across many instances of the same microservice. As Standalone networks expose more APIs and rely on more east-west traffic between microservices, F5’s strengths in L4-7 traffic engineering and service protection become part of the reliability story. Rakuten’s choice signals a priority on resilient traffic handling, predictable customer experiences during peak loads, and a zero-trust mindset for core interfaces that face the public internet.
How a 5G Standalone Core Actually Works
A Standalone core is a set of containerized functions that talk to each other over standard APIs. The key building blocks include:
- Access and Mobility Management Function for registration and mobility decisions. – Session Management Function for establishing and steering user sessions. – User Plane Function for the heavy lifting of packet forwarding. When a failure occurs, orchestration replaces pods, autoscaling adds capacity, and traffic is rebalanced. All of this makes the network more like a living platform than a static appliance farm. For Rakuten, which already approaches mobile like software, Standalone is the natural next step.
What Cloud-Native Means Here
Cloud-native is a phrase that gets overused, so it helps to ground it in specifics: container images for every network function, declarative configuration stored in version control, automated CI/CD pipelines to test new code and deploy it into production, and infrastructure described by code so that environments can be rebuilt on demand. Success depends on disciplined engineering practices. It also means platform teams and network teams work side by side, because cluster health and service health are two halves of the same picture.
Where AI Fits, Beyond the Buzzwords
Operators have used algorithms and rule engines for years. What changes with Standalone is the richness of the data and the control loops you can close. Three areas tend to show the biggest returns:
- Anomaly detection and triage. Models look at per-subscriber latency, attach success rates, paging failures, and packet loss patterns to spot trouble earlier than humans can. – Capacity planning. Forecasting models can map traffic growth down to cell clusters and specific user plane instances, then propose the cheapest way to meet demand. – Closed-loop healing. When a function misbehaves, the system can drain traffic, restart pods, or scale a replacement set while keeping sessions intact. The point is not to remove humans. It is to let engineers spend more time on design and less on repetitive firefighting. Rakuten’s public positioning around AI-assisted operations is consistent with that philosophy, which is to run the network like a modern software service with data-driven guardrails.
What This Means for Consumers
Most people will not notice the change on day one. That is good. The first responsibility of a core migration is that nothing breaks. Over time, customers can expect steadier performance in busy areas, faster session setup, and fewer drops when moving between cells. Voice over New Radio can deliver cleaner calls without shuffling traffic to older networks. For gamers, lower and more consistent latency is the prize. For commuters, the benefit is fewer surprises when thousands of phones wake up in the same train tunnel. The best compliment a Standalone launch can get is that it feels boring. Everything just works, only better.
What This Means for Businesses and Developers
Enterprises care less about raw speed and more about predictable service and integration hooks. A Standalone core gives Rakuten cleaner ways to offer slices or quality tiers to industrial sites, logistics hubs, stadiums, and campuses. The Network Exposure Function can publish APIs that let partners observe service state or request specific characteristics for critical devices. For developers, a richer 5G API surface means smarter applications for vehicles, drones, video analytics, and robotics. With a capable user plane at the edge, application traffic can exit the network closer to where it is used, dropping latency and backhaul cost. Security posture also improves when traffic does not have to traverse the entire core before reaching an application that sits a few kilometers away.
Cost, Efficiency, and the Economics Behind the Decision
Cloud-native architectures are not automatically cheaper. Savings come from standardizing platforms, automating relentlessly, and buying capacity like a software shop instead of a hardware buyer. Done well, operators reduce over-provisioning, scale only what is needed, and shorten incident duration. Vendor diversity also influences cost curves. With multiple suppliers aligned to open interfaces, the operator can encourage competition at the function level without destabilizing the platform. The presence of Cisco, Nokia, and F5 suggests Rakuten wants best-in-class components that can each be upgraded on their own timelines while the overall system stays stable. The expansion of AI-assisted operations is also about labor productivity. If the platform flags problems early, recommends safe actions, and executes routine corrections without human intervention, the operations team can support a larger network with fewer emergency escalations.
Reliability and Security, Explained Simply
Reliability in a Standalone core comes from redundancy and smart failover rather than monolithic boxes with cold spares. Every function runs in multiple instances. If a pod dies, a new one spins up. If a zone fails, traffic is redirected to another zone. The network keeps going because each part is small, replaceable, and protected by constant health checks. Security starts with identity. Every service proves who it is before it talks. Every interface is encrypted. North-south traffic from the public internet meets rate limiters, scrubbing services, and application firewalls. East-west traffic between microservices is controlled by policies in the service mesh. When a vulnerability surfaces, teams patch images and roll out new containers with minimal downtime. F5’s heritage in protecting high-throughput applications and APIs, combined with telecom-grade signaling defenses, is directly relevant as 5G cores expose more programmable surfaces.
How Migrations Usually Happen Without Pain
Operators rarely flip a single switch. Instead, they run the new Standalone core alongside the existing environment and move traffic in phases. Early steps often include:
- Onboarding a small slice of subscribers whose devices are known to work well with Standalone. – Running synthetic traffic to stress test attach, authentication, policy, and handovers. – Steering specific APNs or service classes to the new core while monitoring KPIs in real time. – Rolling back automatically if thresholds are exceeded. The customer experience is steady because the operator has several safe escape hatches. Calls and data sessions continue even if a test fails, and engineers learn quickly without risking the entire base. That engineering discipline, more than any single vendor choice, is what makes or breaks a Standalone launch.
Device Compatibility and Voice
Not every 5G phone uses Standalone features the same way. Two identical models on different firmware can behave differently with the same core. That is why operators work closely with device makers to validate combinations of hardware, software, and radio bands. As Rakuten lights up Standalone services, customers may see new carrier settings updates or firmware recommendations. Voice over New Radio is another area where coordination matters. The network and the device both need full support. Once that is in place, calls stay on 5G, handovers are cleaner, and voice quality improves.
What Sets Rakuten’s Approach Apart
Rakuten’s biggest advantage is cultural. It built its network as code from the start and invested in automation when it was still controversial in telecom circles. This reduces the learning curve for Standalone, because the processes are already software-centric. The new partnerships are less about changing course and more about adding seasoned players to accelerate delivery. The choice to keep working with known partners while adding others indicates a balance between continuity and fresh capability.
Risks and How to Manage Them
No network transition is risk-free. The main hazards include:
- Cloud-native systems produce oceans of telemetry. Without good data engineering, AI models drown in noise. The remedy is to define crisp golden signals and use them to guide both automation and human alerting. Supply and capacity planning. User plane performance can surprise teams when certain packet mixes or encapsulations stress the data path. Synthetic load tests help expose these patterns before customers do. The upside is that each risk is well understood in modern software operations. With the right platform practices, they are manageable.
What To Watch Next
Three signals will show how smoothly the Standalone rollout is going:
- Consistent improvement in attach success rates and session setup times, especially during busy hours. – Expansion of Voice over New Radio as a default for capable devices, with fewer fallbacks to legacy networks. – Real enterprise wins that use slicing, exposure APIs, or edge breakout in ways that were not practical on Non-Standalone architectures. These outcomes take time. The announcement sets expectations, but the real proof is steady execution through the next several quarters.
Practical Takeaways for CIOs and Tech Leads
If you run technology for a large enterprise in Japan, here is how to prepare:
- Check device readiness. Start with one location or one workflow, observe the results, and then scale deliberately. If you are a developer, keep an eye on exposure APIs and edge capabilities. The utility is not just a faster pipe. It is a more programmable network.
Frequently Asked Questions
What is 5G Standalone, in plain English? It is 5G running on a 5G core, without relying on older networks for control. That design enables lower latency, cleaner voice, and more reliable service for complex applications.
Do customers need new SIMs or plans? Most subscribers will not. Operators typically migrate traffic behind the scenes. Some premium features may arrive first on specific plans or devices, but the base experience should simply improve.
Will apps feel faster right away? Some will, especially those that set up lots of short sessions or use real-time media. The bigger gains appear as the network rolls out edge breakout, better scheduling, and Voice over New Radio.
What about security and privacy? Standalone cores introduce stronger default encryption between services and better policy controls. The real safeguard is how well the operator implements zero-trust, traffic inspection for threats, and strict change controls.
How soon will nationwide benefits show up? It is a staged process. The announcement marks the start of deployment. Expect incremental improvements, expanding coverage, and more visible features as testing completes and traffic is steered to the new core.
Conclusion
Rakuten Mobile’s decision to work with Cisco, Nokia, and F5 is a pragmatic next step in its push to operate a software-first mobile network at national scale. Standalone is not a marketing label. It is an architecture that demands discipline, automation, and a new way of running critical infrastructure. The companies involved bring deep experience in core software, traffic management, and cloud operations. Success here will not be defined by a single launch day. It will be measured by quieter dashboards, fewer customer complaints, and the steady introduction of capabilities that were not realistic on older cores. If Rakuten executes the playbook it has outlined, Japanese consumers should notice a network that feels more responsive and resilient, and businesses should find a platform that is finally flexible enough to match the pace of modern applications. The partnerships place Rakuten among a select group of operators that are building the network as a continuously improving cloud service rather than a fixed appliance grid. That is the bigger story to watch over the next year.