1.Adaptive Resource Scheduling

For large-scale heterogeneous resources at the million-level scale, challenges such as high complexity in dependencies and matching between heterogeneous resources, as well as difficulty in ensuring balance after adjustments, must be addressed. To enhance the accuracy of large-scale resource matching and scheduling, auxiliary optimization techniques are employed to generate adaptive intelligent planning and scheduling strategies and models.

2.Unified Orchestration

Based on cloud-network business and operational requirements, customer and service capabilities are translated into service invocation demands for heterogeneous cloud-network resources. Meanwhile, by integrating and abstracting cloud-network resources into a unified service layer, a standardized cloud-network service abstraction is provided, enabling unified management and scheduling of cloud-network resources.

3.Cross-Layer Collaborative Operations and Maintenance

The cross-layer and heterogeneous nature of cloud and network resources increases fault complexity. To address this, an intelligent network fault self-healing engine is required to achieve end-to-end fault localization, root cause analysis, prediction, optimization, and self-healing.

4.Digital Twin

By sensing and collecting information on network and cloud resources as well as their operational status, a dynamic digital mapping of physical cloud-network resources is established. This digital twin model enables real-time simulation and monitoring of cloud-network operations.