China Unicom leverages an autonomous network scheduling platform, gains greater visibility and control
China Unicom built an autonomous, nationwide platform for network scheduling and to open network capabilities via APIs, working with China Information Technology Designing & Consulting Institute Co. The goal was to offer government departments and enterprises innovative products via agile, self-service delivery methods offering unified scheduling, customization, automatic activation, and zero waiting.
The operator faced four major challenges in building the platform. Due to network complexity, the network scheme could not be verified. Configuring resources such as optical fiber and wireless involves verification, allocation and capturing the information. The cost of doing this manually is too high and unstructured.
Service provisioning is a long process with many steps, and done manually is inefficient and error-prone. Overall, there was insufficient automation, end-to-end monitoring and visualization.
more efficient when automatic activation replaces manual data production
a year with network deployed and exposed in minutes not days,
orders per month
faster at less cost
CU-INIC, Deputy General Manager
“Through the application of big data, AI technology and the TM Forum Autonomous Networks Framework, we have realized automatic networking on-demand, automatic resource arrangement, and automatic business activation in the key customer-dedicated lines of services solved the difficult pain points of traditional manual operations – such as low efficiency, high cost and high risk – and shortened the service delivery time from days to minutes. Finally, we improved the efficiency of product delivery and customers’ perception significantly. This is another breakthrough in autonomous networks”.
These challenges were overcome by developing capabilities such as automatic and intelligent networking scheduling, automated provisioning, and full visibility. The partners drew on TM Forum’s autonomous networks guidance to design solutions that include:
• Intent adaptation to capture and analyze customer intentions, then calculate and recommend the most appropriate products.
• Automatic networking uses AI to develop a networking algorithm automatically, such as for a government-private network, which in turn automatically generates a networking scheme.
• Automated set-up by building models in different business scenarios, based on information such as business products to verify and allocate optical paths, equipment, logic and other resources in real-time.
• Intelligent scheduling uses a multi-factor algorithm to calculate routes and multiple path strategies to provide automatic route selection and intelligent cross-domain scheduling based on service level agreements.
• Automatic activation interfaces with network resources to automatically deliver configuration data and activate network exposure. With digital twins of resources, it verifies the simulation of before and after the network exposes capabilities for autonomous, visual control.