DevOps reorgs are consistent challenge to network automation efforts
Operators around the world are rolling out commercial 5G services with a current focus on delivering an enhanced mobile broadband experience to consumers with compatible devices. But, as the technology set matures and the specifications for Release 16 and beyond are fleshed out, the latest generation of cellular will support a massive number of connected things while dynamically providing bespoke network slices connected users, both people and things, to any type of cloud-based application.
To manage no just the sheer complexity of 5G and the use cases it will support, but also help offset the massive capex that comes with a generational network upgrade, network automation is an imperative. But turning over formerly manual processes to software running sophisticated artificial intelligence and machine learning algorithms isn’t exactly a straightforward task.
In a report jointly produced by the MIT Technology Review and Ericsson, the authors see network management as a near-term goal of network automation investments; this has a straight line to reducing operational costs. But beyond just network management, the long-term goal is to build a “self-operating network.” This is particularly important when considered in the context of 5G applications like autonomous vehicles, extended reality, blockchain and edge computing, according to the report authors.
“The end goal is to create a ‘self-operating network’—one in which all functions are completely programmable, without the need for manual configuration and troubleshooting.”
In practical application, operators are tapping open source software and multi-cloud strategies to orchestrate and manage widely-variable types of network workloads. AT&T regularly updates on its core virtualization efforts built and Verizon recently used Kubernetes to orchestrate a container-based, cloud-native evolved packet core trial.
In a blog post, Ericsson’s Ana Freitas called out two key challenges facing operator network automation efforts, quoted below:
- Face up to disruption –Structural changes are necessary to gain the benefits of automation. Eventually, legacy systems will only add complexity and cost to network operators.
- People challenges are far tougher than the technology ones – Operators are making structural changes to capitalize on automation. These often involve merging or redistributing responsibilities across network and IT teams. However, retraining existing staff and trying to instill DevOps principles are by far the toughest challenges.
