The digital ecosystem is expanding and 5G technology, including 5G slicing, is becoming increasingly pervasive, delivering the flexibility, scale, and openness to extract further value from the network. As a result, communications service providers (CSPs) are forming more B2B partnerships across industries; delivering services according to business intent will be foundational to the success of these relationships. This energizes CSPs to have a completely new view towards autonomous operations, needing a guaranteed closed-loop environment that elevates operations to service level and ensures adherence to service level agreements (SLAs) in faster-than-ever changing network conditions. Modern service and network assurance helps CSPs meet this critical objective of achieving autonomous assurance with business context.

Intelligent automation is the key to autonomous assurance 

Autonomous operations push the boundaries of assurance into data-driven decision systems and closed-loop management. That entails a unified view through observability across multiverse (layers, domains, vendors, clouds, and technologies), AIOps-driven analytics, and automation.  

Observability deals with dynamic service, network and resource topology, modeling, and stitching all information together. AIOps analytics brings all the intelligence for data-driven decisions and predictive assurance, with ML capabilities to detect anomalous behaviors and assess impact and causalities to drive the automation layer. AIOps automation delivers the closed-loop control for preventive SLA adherence and drives automated actions for optimal service experience, including healing, optimization, and simulation scenarios.  

Nokia addresses the full autonomous operations stack with Digital Operations Center which is a modular solution consisting of two independent and closely integrated products, Orchestration Center and Assurance Center, underpinned by Unified Inventory. Autonomous assurance, as depicted in the picture above, is achieved by Assurance Center, which is an advanced tool designed to help CSPs apply embedded knowledge across detection, decision, and remediation, using machine learning (ML) to increase the value and accuracy of actionable data used in an operational context. This includes many use cases to address different areas, such as: 

• Closed-loop with orchestration to connect design and operate phases for end-to-end service lifecycle management and faster time-to-value applying ML techniques to identify and resolve anomalies for non-alarming issues before they generate service and customer impact events
• Using data sources and feeds for decision-making purposes from other systems such as NWDAF (e.g. Nokia AVA NWDAF) and MLOps Frameworks (e.g. AVA Open Analytics) 
• Foresight on network and service quality for preventive SLA management
• Operational intelligence using topology correlation, automated root cause ad ML outcomes to drive automated operations
• Service intelligence with end-to-end service stitching to monitor service experience, assess and prioritize by impact 
• Automated and cross-domain optimization of network functions to ensure end-to-end performance within SLA targets. 
• Preventive service quality management by applying ML techniques to identify and resolve anomalies for non-alarming issues before they generate service and customer impacting events. 

How does Assurance Center address top CSP challenges?

With Assurance Center Nokia is committed to providing our customers a reliable and efficient solution to their multiple network and service challenges. 

Assurance Center reduces network noise and focuses on incidents based on priority and severity while applying techniques that allow early identification of faulty situations, preventing network issues that might lead to outages. The observability capabilities across services, networks, and infrastructure combined with ML-based enablers for correlation, forecasting, causality detection, and alarm pattern recognition ensures that anomalies and service-affecting degradations are quickly identified and acted on. Experience from the field indicates that the number of events to be handled can be reduced by 95 percent to 98 percent by applying those techniques above. 

At the service level, predictive handling of service issues helps CSPs avoid penalties and ensure adherence to agreed SLA targets. While monitoring services and slices end-to-end, Assurance Center can proactively identify SLA breach risks and trigger actions to act at the resource level to dynamically change performance targets across different domains and optimize end-to-end SLA targets. Additionally, Assurance Center is optimized for 5G and 5G slicing assurance, including NSMF and NSSMF functions (radio and core) providing automated resolution of anomalies and service degradations in an automated closed loop. Pre-built KPI knowledge is used together with capabilities to forecast anomalies and dynamic topology correlation to conduct accurate service impact and root cause analysis. Based on the experience from the field, on a typical CSP network, 40% of the incidents can be automatically resolved without any human intervention. 

Conclusion

On the journey towards autonomous operations, assurance is bringing observability, ML-powered analytics, and automation together to deliver AIOps capabilities. The focus is on adaptive operations through network, service, and 5G slice foresight on performance and SLA progression with advanced mechanisms to continuously detect anomalies and optimize root cause and service experience. The main benefits are fixing issues faster, reducing noise, and preventing issues before they happen. The outcome will not only lower operational costs but improve resolution time. This is critical to improve the effectiveness of operations and meet business targets agreed upon with B2B partners.