Leveraging Generative AI (GenAI) for service and network operations is not simply a technological advance but a transformation that delivers continuous improvements by extracting and creating new knowledge. Let’s explore how GenAI powers full autonomous operations of networks that sense, think and act.

The journey towards full autonomous operations

Communication Service Providers (CSPs), neutral hosts and enterprises are pushing the boundaries of managing next-generation networks towards full autonomous operations, supporting the entire lifecycle with closed-loop automation across services and domains.

We have now reached a stage where AI/ML is an essential reality when managing networks and AIOps has been embraced to deliver AI-led closed-loop control. AIOps approaches are highly effective in solving well-defined problems such as detection of anomalous behaviors, understanding causality between events, identifying patterns or predicting behaviors.

As discussed in a recent blog post, GenAI provides a new and unique approach to AI, shifting towards highly flexible and adaptive systems that create new knowledge. This new approach augments the capability of systems to behave autonomously and to assist service operations. GenAI can help to streamline complex tasks that traditional AI does not address, such as summarizing knowledge, simplifying human-machine interactions, generating content, making decisions and providing recommendations based on unstructured data.

Networks that sense, think and act

Nokia believes that networks of the future need to sense, think and act. Enabling autonomous operations plays a pivotal role in this evolution by sensing with observability, thinking with AI and acting with closed-loop automation as shown in the diagram below.

The “sense layer” collects different data sets across domains, vendors, technologies and layers to provide the visibility required for better decision-making. The “think layer” creates knowledge by performing analysis based on context and target to come up with recommendations and take decisions. The “act layer” performs actions to ensure effectiveness of automation processes.

As shown in the diagram below, GenAI is a part of the “think layer” and can create new knowledge which can then be leveraged in all three layers. For example, GenAI-generated collection adapters for the “sense layer” could be used for new incoming data, while new automation agents created by GenAI could be leveraged in the “act layer”.

Enabling new autonomous operations use cases with GenAI

CSPs are on a journey towards full autonomous operations: network digital twins, AIOps, intent-based orchestration, and now GenAI are key capabilities to achieve this goal. As shown in the diagram below, these capabilities complement each other, supporting the following objectives:

• AIOps-led assurance for AI-driven closed-loop control.

• Digital twin-first approach to pre-verify outcomes for planning, operations, maintenance, care and innovation purposes.

• Intent-based orchestration providing focus on business outcomes.

• Self-evolving systems with GenAI to increase autonomy of operations from intent to execution.

GenAI can be applied to dynamically create new data and knowledge. This new data can then be used by the system to self-evolve and to allow humans to be involved as supervisors rather than doers. In general, GenAI helps to increase the autonomy of next-gen networks, while working together with other autonomous operations enablers. For example, AIOps use cases can benefit from GenAI capabilities to summarize data and to create dynamic sets of information for predictions.

But most importantly, GenAI opens the door for a wide range of completely new use cases including the ones outlined below:

• Co-pilot and user interaction use cases: This is about assisting humans with different tasks, either by guiding them through the available information to simplify analysis or by auto-generating content that would traditionally be provided by users. This can help with incident analysis and reporting, service design order processes, to name just a few examples. 

• Towards autonomous operations maturity level 5: GenAI is one of the key capabilities that enables systems to operate with a higher degree of autonomy towards TM Forum autonomous operations maturity level 5. To reach that level, additional system capabilities are required for intent, awareness, analysis, decision and execution. For example, the system can negotiate and resolve intents, or create new artifacts for AIOps-led processes and closed-loop management without human intervention.

• Autonomous system capability augmentation: In addition to having the system operate in a fully autonomous loop, there are also capabilities that increase system efficiency and allow exploring alternative options. For example, the system can assist users to create custom content and configurations to optimize the closed-loop process for all stages of the design, deploy and assure service lifecycle. 

Conclusion

As evidenced by our award-winning Digital Operations Center solution, Nokia is invested in making the autonomous networks concept a reality. This is enabled by capabilities that sense the environment, think to adapt the behavior and act accordingly with no human intervention. Due to the inherent complexity of managing networks, different enablers are required to work together to reach that state. GenAI is taking a prominent role due to its flexible and adaptive nature to understand unstructured data, context and dynamically create new content and knowledge.

To learn more or to request a demonstration please visit our website.