AI and ML embedded on all network layers is the gamechanger for network experience

The complexity and density of data traversing the broadband network is increasing as more advanced technologies and applications are being adopted within the home. It is essential that service providers can maintain customer loyalty and provide an enhanced Quality of Experience (QoE) while avoiding rising operations and management (O&M) costs. Service providers need to leverage the intelligence of AI and ML in the cloud to digitalize and monitor vital data in their physical networks was the message expressed by Aihua Guo, Principal Architect, Futurewei.

Moving the broadband network from a mix of infrastructure to an all-fiber network that is deployed end-to-end, connecting the end-user all the way to the cloud, can provide a greater overall QoE for end-users. An all-fiber network can deliver a broadened bandwidth spectrum and premium services with a reliable performance which is difficult to achieve with older technologies, according to Guo.

But with the number of subscribers on the network growing significantly and therefore an exponential amount of data generated, human-aided O&M becomes impractical and related costs increase. With different management systems in the operator’s network, the amount of data is becoming so huge that it is difficult to manage in traditional networks. If operators are unable to manage applications, user experience, and identify issues on the network efficiently, they run the risk of losing subscribers.

There has been a paradigm shift in the connected home sector to accommodate for applications such as live streaming and Augmented or Virtual Reality that require real-time connectivity and low-latency experience. With greater use of and reliance on these applications in the home, if incidents are not self-identified and only discovered due to complaints, it can lead to further customer churn.

Guo said that for the broadband network to become more self-driving, it needs help from intelligence from closed-loop automation with AI and ML embedded on all layers of the network in each network domain. Closed-loop AI is a predictive analytics software that uses ML to predict the future. This can provide the likes of resource management provisioning without manual intervention.

Guo claimed “A cloud-based platform that is digitalized and has AI capabilities can provide a solid foundation for the network and its applications. To fully unleash the power of AI and closed-loop automation, intelligence can be deployed in the cloud and distributed across computing-powered access devices on the network to enable intelligent coordination and data collection between them.

“It can essentially build a holistic view of the physical network with a digital twin that has an existing copy of all its data. The digital twin can therefore run all the services, applications, and experience on the network and present real-time solutions to the service provider to identify low QoE and optimize performance.”

During the presentation Guo discussed an Automated Intelligence Management (AIM) framework use case of Intelligent Wi-Fi Interference Control. He highlighted that dense Optical Network Unit (ONU) deployments with Wi-Fi access points in a small area serving multiple rooms in a home, can result in severe signal interference and degraded application experience. An AI model can generate real-time data on the status of all Wi-Fi devices on the network and analyse the interference hotspots. It also creates optimal Wi-Fi configurations which is pushed to the ONU to alter its configuration.

The evolution in user applications requires enhanced network automation and intelligence to adapt to the changing landscape. A network empowered by AI and ML facilitates automated O&M that can provide greater QoE and reduce costs, concluded Guo.