ABSTRACT

This chapter considers some implications of applying artificial intelligence (AI) and particularly machine learning (ML) to cloud-based communication networks providing complex computing services. Developing, testing, deploying, and securing AI-based communication services for cloud-based autonomous networks is challenging and complex. Cloud technology, including containers, cloud natives, and microservices, enables reusability of applications, reduced time to market, optimal use of infrastructure, and detailed adaptation to users' needs; however, the administration complexity calls for AI. One microservice may apply for a service from any other microservice via the cloud interprocess communication infrastructure. Virtual machines became a very useful, if not essential, cloud infrastructure tool, enabling flexibility where an application may run anywhere, irrespective of the underlying hardware and Operating System (OS). Network function virtualization (NFV) promotes decomposition of the functionality of a network node as much as possible. In an NFV-based network, services that can be implemented in software are executed over generic cloud hardware as virtual network functions (VNFs).