Transforming network service assurance with real-time AI insights

This article originally appeared in Vanilla+ in August 2023. Link is here:https://www.vanillaplus.com/2023/08/16/82798-transforming-network-service-assurance-with-real-time-ai-insights/

Communications service providers (CSPs) face rising demands for new business-focused applications, improved service quality, and real-time resolution of issues. Alongside these challenges lie increasing complexities whereby operators need to monitor and troubleshoot innovative solutions, including 5G and the internet of things (IoT), without onboarding additional resources, says Matthew Twomey, product marketing lead, Anritsu.

In the dynamic era of 5G and beyond, network service assurance is undergoing a transformative shift, posing new challenges for operators to assess and overcome. While tackling the myriad of challenges is an ongoing effort, the key to continued network success lies in real-time visibility of subscriber/device issues and the seamless application of artificial intelligence (AI).

As the complexities of network technologies continue to intensify, AI will increasingly take centre stage in the transformation of network service assurance. For instance, AI-driven analysis ensures faster issue resolution, improved network efficiency, and enhanced customer experiences. In this article, I will explore network technologies’ rapid evolution, increasing complexities, and innovative approaches to overcome challenges. I will also look at service assurance from a fulfilment and network perspective.

Stay ahead of the game with a cutting-edge approach

The future success of network service assurance relies on several initiatives real-time visibility of issues, the application of AI in real time, and cloud-native architecture which requires new and innovative tactics.

• Real-time visibility of subscriber issues: Network service assurance traditionally focused on monitoring network functions and infrastructure to ensure seamless connectivity. However, increasing network complexity and rising customer expectations are creating a shift to real-time visibility of subscriber issues. This shift requires the ability to understand the customer experience, detect and resolve service issues in real time, and proactively address potential issues before they negatively affect user experiences.
• The power of AI: With the increasing complexities of networks, manual monitoring and intervention have become less efficient and less effective, and this is where AI plays a pivotal role. Poised to revolutionise network service assurance, its real-time application can analyse vast amounts of data, predict potential issues, and optimise network performance, empowering service providers to proactively address anomalies. Real-time AI-driven analysis looks at subscriber-level details, allowing operators to detect and resolve issues faster, improve network efficiency, and enhance the overall customer experience.
• Cloud-native architecture: Essential to keeping pace with the increasing demands of businesses and subscribers, transitioning to cloud-native architecture while supporting technologies like private networks is quickly becoming table stakes. However, this transition requires operators to continually adapt to evolving 3GPP standards. The continuous redefining of network specifications has introduced escalating complexity for operational teams, requiring them to adapt to new tools, protocols, and processes. This has necessitated a continuous learning process to ensure teams have the skills and knowledge to manage and assure cloud-native networks effectively. In order to guarantee a smooth transition to an automation-driven operational team, AI-assisted assurance systems are now becoming a prerequisite.

Understanding the Game

Service assurance can be approached from two perspectives: fulfilment and network. From a fulfilment perspective, service assurance focuses on the end-to-end process of delivering services to subscribers, including order management, provisioning, and activation. Its emphasis is on seamless service delivery to meet customer expectations but not on the running of those systems. This service assurance also needs real-time data, AI/ML, and is cloud-native but does not use real-time customer traffic of the services.

Service assurance from a network perspective concentrates on monitoring and optimising the health and performance of the network infrastructure to deliver a positive customer experience. While individually each assurance perspective is important, both are critical for a comprehensive approach to service assurance. Operators that address service assurance from both angles can better manage network resources, provide high-quality services, and deliver exceptional customer experiences.

Data when and what

In the competitive world of telecoms, ensuring customer experience and network service is critical. It’s not only using data but knowing which data to use. Data timeliness or when you have the data can be viewed as split into three buckets: real-time (within a second), fast-time (within a minute), and key performance indicator (KPI)-time (every 5/15 minutes or longer).

Real-time data, coming from the network, is used for immediate tasks like spotting network faults and managing traffic, using tools like AI for quick decisions. Fast-time data, while not instant, is used for tasks like checking the quality of service and looking into how services are performing for subscribers from their perspective. KPI-time data is used for longer-term tasks and offers a detailed look at the network’s overall performance, using deeper analytics for insights. Operators must understand what data is needed, when it’s needed, and what impact it has before they can start their journey to using AI/ML for automated networks.

Overcoming future network challenges

To understand subscribers’ experiences, service providers have relied on the data they collect from between the nodes in what were standard interfaces. 5G networks have upgrades in security built into the network core, making data acquisition of customers’ experiences more challenging. In effect, the data is encrypted, and the network software vendors must now facilitate this data acquisition via the introduction of vTaps (virtual taps). Without the data to understand the network’s impact on the customers’ experience, operators cannot take the necessary corrective actions manually or automatically putting them at a distinct disadvantage.

The success of future network service assurance requires a paradigm shift from simple real-time visibility of network functions to real-time visibility of subscriber issues. It is only through this approach that a more customer-centric model will become reality, opening the door to enhancing the subscriber experience and driving automation. Already critical to network success, AI’s real-time application will continue to take on an increasingly crucial role in addressing the complexities of network management and enabling seamless and proactive troubleshooting.

As the industry continues its transition to cloud-native architecture and adapts to ever-evolving 3GPP standards, operational teams will find themselves navigating escalating complexities. With AI playing a central role in delivering real-time insights, operational teams will shift from manual operations to become an automation-driven operational team, making end-to-end customer and network visibility a reality.