AI-driven automation is going to change the way we run our networks end-to-end, from the access in the RAN all the way to the core, from the edge infrastructure to the cloud. To get there, we need the right silicon and platforms. Are we there yet?
In this Sparring Partners, Cristina Rodriguez at Intel and Monica Paolini at Senza Fili talked about the network transformation that AI and automation enable, and how the industry ecosystem is already working on multiple use cases to make wireless networks more efficient.
– Are we ready to fully embark on AI-driven automation?
– Do we need automation more or less as we move to cloud-native, software-defined, disaggregated and virtualized networks?
– Is AI going to accelerate the deployment of open and virtualized RAN?
– What steps and best practices will take us to AI-driven automation and, eventually, to autonomous networks?
Watch other Sparring Partners or register for upcoming ones at
Timestamps:
00:04 AI applications across the network from edge to cloud
04:11 AI applications in wireless networks, including power management, predictive maintenance, and network slicing, with examples of successful demonstrations and potential for AI. Intel projects with Vodafone, Deutsche Telekom and Ericsson, demonstrating Xeon’s capabilities
14:36 Using AI in telecom networks to improve efficiency, reliability, and security
21:22 Using AI to improve network efficiency and predict future issues
25:08 AI in 5G networks used for network slicing and conflict resolution, to lower TCO, ensure SLSs, reduce costs, and improve efficiency. Role of AI to manage complexity
33:05 AI in Open RAN increases efficiency and accelerates innovation
37:23 Intel’s role in providing AI solutions, Xeon, AI built-in acceleration, reduction in power consumption, scalable architectures, importance of software
44:41 AI for video processing, power consumption, and deployment options (cloud or edge?)
49:02 AI at the edge for latency-sensitive applications
54:02 Trust in AI models for network optimization
58:24 Data sources for AI models and long-term goals
[ad_2]
source
