RAN Datasets Must Evolve to meet AI/ML Objectives

Using standard datasets for Artificial Intelligence (AI) and Machine Learning (ML) analysis of Next Generation Radio Access Networks (RAN) will not solve many of the critical RAN capacity and coverage issues or automated intelligence goals due to their complex objectives. Legacy signaling datasets can no longer simply be correlated to resolve performance issues and service objectives for these advanced 4G/5G networks.

Communications Service Providers (CSPs) know that the only path forward to next-generation AI/ML automation intelligence framework is to apply stateful analysis of raw RAN datasets and all critical emerging network variables prior to feeding AI/ML frameworks. And this path requires a journey of discovery to analyze data at the source.

With 30+ years of experience working with data at NETSCOUT, we know data secrets and we are helping CSPs succeed with their next generation 4G/5G networks.

Join NETSCOUT’s RAN expert, Robert Froelich, Sr. Director of Product Management, in this webinar, to learn about these four critical issues in 4G and 5G Non-Standalone (NSA) and Standalone (SA) networks:

• The impact of poorly configured neighboring cells on service and subscriber experience
• Why single call analysis is no longer viable to optimize carrier aggregation
• Why optimizing intra-cell elements for capacity is important
• Why AI/ML algorithms need network source, problem, or use case-specific datasets

The 5G core is essential to enable standalone (SA) operation in the 5G system. Operators are currently designing and deploying new core network architectures using cloud native technologies and operating models. This investment will have deep and broad implications for operator services and the wider mobile ecosystem.

This webinar presents the key findings from the second annual Heavy Reading Cloud Native 5G Core Operator Survey. Heavy Reading’s senior principal analyst for mobile networks will lead a panel to discuss and analyse the results of the 2022 survey, with guest speakers from AMD, F5, Oracle and Red Hat contributing expert insight and opinion.

Among the topics to be discussed include:

• The impact of 5G core on services and customer experience
• 4G to 5G core migration strategies and common 4G/5G core networks
• Cloud infrastructure choices for 5G core (IaaS vs PaaS vs vendor stack)
• CI/CD/CT tools and operations
• The outlook for edge services and distributed 5G core

In the future, individuals, households, and industries will have higher requirements for digital infrastructure.

For individuals, immersive services like XR and holographic communication are maturing, and connectivity experience is set to increase from 1 Gbit/s to 10 Gbit/s. Mobile DOU will surge from today's 15 GB to 100 GB. Requirements for latency and ubiquitous connectivity will also increase. For households, the demands generated by advanced services like 24K 3D VR games and holographic education and meetings are creating a full-fiber, 10 Gbit/s era. For industries, digital transformation has now entered the fast lane. Industrial-grade applications such as smart manufacturing and power grid dispatching are raising diversified requirements for connections, quality, and sensing, while also triggering explosive growth in demand for computing power and storage.

We must continue to work hard if we hope to reach an intelligent world. The 5.5G era is an important milestone on this path – one we cannot miss. In the 5.5G era, we will need ubiquitous 10 Gbit/s experience; intelligent and high-quality compute scheduling; highly autonomous L4 networks; Cloud Native 2.0 services for enterprises, and a 10-fold increase of computing effectiveness, storage, and infrastructure energy efficiency.