The telecommunications industry stands at a critical juncture. The promise of 5G, IoT and edge computing has unleashed a data deluge, yet telcos worldwide are facing an existential crisis. The demand for always-on, zero-downtime connectivity is colliding with immense pressure to cut costs.
In Europe, for example, many telcos face a mandate to reduce operational expenses by as much as 30% through automation and AI. This is a double whammy: operate more efficiently with a torrent of new data while the current operational model—built on manual interventions across siloed systems — is fundamentally broken. Network management is often a reactive firefighting exercise, focused on hardware availability rather than proactive customer experience.
The future vision for 6G is built on the concept of autonomous networking, where AI makes real-time, self-correcting changes. Those who can’t achieve this will be priced out of the market.
This isn't just about efficiency; I believe it's about survival. The key to navigating this crisis is the network digital twin.
Building the network digital twin
The industry understands that the path forward lies in a true network digital twin — a real-time, complete representation of both the physical infrastructure and virtual network layers. Unlike simple network monitoring, a digital twin goes beyond structured operational data by incorporating unstructured data streams, such as maintenance tickets, resolution logs and external events. This holistic intelligence is what enables autonomous decision-making.
The business benefits are significant. Beyond the crucial reduction in opex, a digital twin enables telcos to predict and prevent network issues before customers are even aware of them, dramatically reducing churn and paving the way for new revenue streams.
By simulating network changes in a digital twin before deployment, telcos can ensure stability and performance, moving from a reactive to a proactive operational model. But for this vision to become a reality, telcos must first solve a fundamental problem.
The architecture gap that causes new concerns
As telcos embrace the digital twin, a hidden obstacle has emerged: the underlying data architecture. The core issue is the databases themselves, as they force an impossible trade-off. That is, to create a complete view of the network, you need two fundamentally different data models that traditional databases can't handle together.
A network's complete state requires two different, and often incompatible, data models. Network topology needs a graph database to model the complex relationships and connections between devices. Meanwhile, operational telemetry — encompassing performance metrics, time-series data and status updates — requires a relational structure for efficient analysis and querying.
Current solutions force operators to either choose one or, more commonly, maintain multiple separate systems. This fragmented approach defeats the entire purpose of a unified digital twin.
Think of a network anomaly as a medical diagnosis. To find the root cause, you can't just look at one symptom. You need to combine the patient’s physical connections (graph data), their vitals over time (relational data) and their medical history (unstructured reports) to get the full picture.
Unfortunately, traditional architectures scatter this information across multiple systems, making real-time correlation impossible. By the time you've manually pieced together the full picture, the network has likely already failed. This architectural fragmentation is the hidden obstacle blocking the path to true network intelligence.
Blueprint for the true digital twin database
To truly enable network digital twins, a new data foundation is needed—one that overcomes the limitations of traditional database architectures. There are four core capabilities that I believe this system must deliver:
Unified, instant insight: the ability to model complex network relationships alongside time-series operational data in a single system. This eliminates the need for separate databases and the complex, slow process of synchronizing data between them, allowing for real-time diagnosis across all dimensions of the network.
Planetary-scale consistency: strong global consistency across geographic regions is non-negotiable. Autonomous decisions must be based on a single source of truth, not rough approximations that can lead to cascading failures or major outages.
Multi-modal data handling: native support for all types of data — including performance metrics, unstructured incident reports and external correlation sources like weather data, public event schedules and construction plans — within the same real-time framework. This allows for a richer, more contextual understanding of the network's state.
Real-time, low-latency processing: The system must process massive data streams with sub-millisecond precision. This is essential for autonomous operations, where even a slight delay in data analysis could lead to a network outage or service degradation.
This is the blueprint for a database that doesn't just store data — it unlocks network intelligence. It's a foundation designed to handle the complexity and scale of modern networks, unifying previously fragmented data.
From database compromise to network intelligence
True autonomous networking requires a fundamental shift in data architecture, one that breaks free from the traditional need to compromise between graph and relational systems. Forward-thinking telcos are already making this shift, recognizing that only cloud-native solutions can deliver the massive scale and instant data correlation that autonomous networks demand.
Before investing in more AI or orchestration tools, telco leaders must first evaluate whether their data architecture can actually support the digital twin intelligence that autonomous operations require. The database isn't just infrastructure — it's the foundational blueprint for the industry's transformation.
Angelo Libertucci is global head of telecom at Google Cloud.
Op-eds from industry experts, analysts or our editorial staff are opinion pieces that do not represent the opinions of Fierce Network.