Artificial Intelligence For Media Operations Why AI algorithms will become a must for every network management system

摘要

Artificial intelligence already has a big impact on a lot of different segments in the media industry. With the broadcast industry moving to all-IP and data center deployments those environments are more agile and complex than ever before. With fast-evolving technologies and cycles, ever more mission critical systems and constantly changing operational practices and devops style operations, traditional network management systems (NMS) and their paradigms don't fit the bill anymore. This paper explores why machine-learning algorithms must find their way into network monitoring and management solutions to orchestrate a modern media data center dynamically and in a proactive way. The foundation for an AI-driven NMS platform is a solid big data storage architecture; a lack of profound data and data hygiene is often one of the biggest obstacles to successfully deploy big data projects. As in today's all-IP environments nothing is static anymore, an AI entity must be highly intelligent. Unsupervised learning is key to automatically adapt to changing environments. Augmented operation and a zero-configuration and zero-maintenance approach will be crucial to successfully deploy the right management strategy.