TOSCA Service Instance Templates for Low-Latency 5G Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current TOSCA-based orchestration in 5G networks experiences significant latency due to the time-consuming translation of service templates to service instances, which is unacceptable for zero-touch automation.
Innovation Solution
A TOSCA-based orchestrator accesses a template catalog storing network service instance templates, allowing direct deployment of these instances to reduce latency by avoiding the need for real-time data gathering and template translation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If TOSCA-based orchestration uses service templates stored in a template catalog, then the system maintains standardized orchestration processes, but the translation from service template to service instance introduces significant latency of minutes to dozens of minutes
Solution Approach 1:
The patent pre-generates service instance deployment packages at design time by translating service templates into executable deployment artifacts and storing them in a deployment package repository. This preliminary action eliminates the need for real-time template translation during runtime, reducing latency from minutes to seconds while maintaining standardized orchestration through the pre-defined deployment packages.
Solution Approach 2:
The system performs policy evaluation and inventory data gathering in advance during the design phase, embedding these results into the pre-generated deployment packages. This eliminates the need for time-consuming policy lookups and inventory queries during runtime LCM operations, significantly reducing the latency while ensuring consistent policy enforcement.
2Manufacturing precision
If the orchestrator performs multiple time-consuming operations to gather instance data (policy, inventory lookup, etc.), then the system ensures accurate and complete service instance deployment, but the process introduces unacceptable latency for 5G network automation
Solution Approach 1:
The system performs all necessary data gathering operations including policy evaluation, inventory lookup, and parameter validation during the design phase when service templates are translated into deployment packages. These pre-computed results are embedded into the deployment packages, ensuring deployment accuracy is maintained while eliminating time-consuming operations during runtime LCM executions.
Solution Approach 2:
The patent creates copies of service templates in the form of pre-generated deployment packages that contain all necessary configuration data, policy evaluations, and inventory information. These deployment package copies encapsulate the essential deployment information without requiring access to the original template catalog or external data sources during runtime, thus maintaining accuracy while reducing latency.
3Adaptability or versatility
If LCM operations include multiple network configuration operations in addition to cloud infrastructure manipulation, then the system achieves comprehensive service management, but the complexity of coordinating these operations increases latency
Solution Approach 1:
The patent merges multiple separate operations including service template translation, policy evaluation, inventory data gathering, and deployment package generation into a single pre-processing phase. By combining these operations that were previously executed sequentially at runtime into one upfront preparation step, the system maintains comprehensive service management capability while dramatically reducing the latency of LCM operations to seconds.
Solution Approach 2:
The system performs all complex coordination tasks including network configuration parameter preparation, policy validation, and infrastructure resource allocation in advance during the design phase. These preliminary actions result in self-contained deployment packages that can be executed rapidly at runtime without requiring complex coordination during LCM operations, thus maintaining versatility while reducing latency.
Data Source
AI summary
As described herein, a system, method, and computer program are provided for a TOSCA modeling optimization for 5G network orchestration. A template catalog storing one or more network service instance templates is accessed, by a TOSCA-based orchestrator in a 5G network. A network service instance is deployed to the 5G network from the one or more network service instance templates, by the TOSCA-based orchestrator.