Distributed NFV Orchestration for 5G Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing Network Functions Virtualization (NFV) platform faces challenges in scaling to support the deployment of virtual network functions (VNFs) across thousands or tens of thousands of sites, particularly in meeting the speed and latency requirements of 5G networks.
Innovation Solution
A distributed orchestration framework is employed, allowing VNFs to be deployed across different types of cloud computing data centers under the control of a central orchestrator, using static inventory data to select the appropriate data center for deployment and issuing generic commands that can be translated into API calls recognizable by various cloud computing management software, enabling seamless deployment across multiple sites.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a centralized NFV platform is used to manage VNF deployment, then control and coordination are simplified, but the system cannot scale to support thousands or tens of thousands of sites
Solution Approach 1:
The patent divides the centralized NFV platform into a hierarchical structure with a central NFV orchestrator for strategic coordination and local NFV managers for tactical deployment execution. This segmentation allows the system to scale to thousands of sites while maintaining manageable complexity through clear division of responsibilities between central and local entities.
Solution Approach 2:
The patent introduces local NFV managers as intermediary components between the central orchestrator and VNF deployment infrastructure at each site. These intermediaries translate centralized policies into local deployment actions, enabling scalability without requiring direct central control over every deployment detail, thus reducing overall system complexity.
2Speed
If VNFs are deployed closer to end users to meet 5G speed and latency requirements, then network performance improves, but the number of deployment sites increases dramatically
Solution Approach 1:
The patent implements local NFV managers at each deployment site that are specifically configured for that location's requirements. Each local manager handles VNF deployment decisions tailored to that site's network conditions, user base, and infrastructure capabilities, enabling optimized performance at each location while managing the complexity of deploying across many sites through standardized local patterns.
3Adaptability or versatility
If different types of VIM software are used across data centers, then flexibility and vendor diversity are improved, but compatibility and interoperability become difficult
Solution Approach 1:
The patent designs the NFV orchestrator with universal interfaces that can communicate with multiple types of VIM software through standardized protocols. The orchestrator maintains a catalog of supported VIM types and automatically selects appropriate VIMs for deployment, enabling multi-vendor flexibility while simplifying deployment processes through unified management abstraction.
Solution Approach 2:
The local NFV managers serve as intermediaries that translate between the universal orchestrator interface and the specific VIM interface required by each data center. This intermediary layer handles the complexity of VIM-specific protocols and APIs, allowing the system to deploy across diverse infrastructure types without requiring direct knowledge of each VIM's implementation details.
Data Source
AI summary
A method of deploying a network service across multiple data centers, each having a cloud management server running a cloud computing management software to provision virtual infrastructure resources thereof for a first tenant among a plurality of tenants, includes maintaining for each data center static inventory data that indicate virtual infrastructure resources that are available thereat to the first tenant, identifying, in response to a network service request for the first tenant, a virtual network function associated with the network service, generating commands to deploy the virtual network function based on a descriptor of the virtual network function, selecting one of the data centers in which the virtual network function is to be deployed based on the descriptor of the virtual network function and the static inventory data of each data center, and issuing the commands to the selected data center to deploy the virtual network function.