Читать книгу Artificial Intelligent Techniques for Wireless Communication and Networking - Группа авторов - Страница 45

A 5G networking feature known as network slicing will be augmented by AIOps tooling. This allows many virtual networks to be run over one physical connection by 5G networks. AIOps tooling will allow proactive and dynamic supply of individual wireless service quality levels for many customer categories and edge device groups by using this virtualized resource allocation capacity. AIOps will also be required to increase dynamic 5G RF-channel allocation.

5G has smaller cells than 4G at each edge unit, reuses wavelengths more thoroughly, and must re-target “beam formed” base station phase-array millimeter-wave antennas continuously. In order to secure service quality, 5G base stations instantly predict as well as provide each user with the best wireless router. They do this because the problems associated with 5G mm waves moving through walls and other hard surface areas are constantly being tackled. In order to perform these measurements in real time over constantly evolving local wireless loops, a securely shuttered real-time analysis is necessary.

All this AI inside the 5G network, for example, would have a support system for data storage. We would expect to see regularly develop in 5G networks specialist data lakes, autoML applications, DevOps databases and other critical operational architectures to ensure that the best AI models are implemented in real time. This data/model governance will be introduced through cloud-to-edge architectures that fit into complex public/private federal environments typical of 5G [17, 21].