VNF Allocation Engine Density-Based Grouping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Virtual Network Function (VNF) clustering techniques often result in non-optimal software clusters due to the removal of edge VNFs, leading to suboptimal communication and increased latency across separated VNFs.
Innovation Solution
An information handling system with a VNF allocation engine that determines physical and virtual groups based on traffic ratios and node densities, prioritizing bandwidth conservation and power efficiency by merging VNFs with similar functions and provisioning them on the densest physical groups, typically within the same server or rack.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional cluster placement techniques are used to group VNFs into software clusters and provision them on hardware clusters, then the communication distance between VNFs is reduced and the number of computing devices is minimized, but edge VNFs are removed from the system and VNFs that need to communicate are separated into different clusters
Solution Approach 1:
The patent changes the optimization parameter from bandwidth-based edge VNF removal to density-based virtual group formation. By using the ratio of traffic leaving a group to the number of VNFs in the group as the optimization criterion, the system identifies densely-connected virtual groups that maintain communication integrity while still achieving physical consolidation benefits.
Solution Approach 2:
The patent segments the virtual topology into multiple virtual groups based on density criteria, allowing different virtual groups to be provisioned on different physical groups. This segmentation maintains the connectivity of VNFs within each virtual group while enabling efficient physical resource utilization across the infrastructure.
2Productivity
If edge VNFs are removed to break up primary software clusters into smaller secondary clusters, then bin packing algorithms can maximize traffic within VNF systems, but too many edge VNFs are removed when multiple edge VNFs transmit the same amount of traffic and non-optimal clusters are created
Solution Approach 1:
The patent replaces the bandwidth-based parameter with a density-based parameter (ratio of traffic leaving the group to number of VNFs in the group). This parameter change provides a more balanced criterion that considers both traffic patterns and cluster composition, avoiding the arbitrary removal of edge VNFs that transmit equal amounts of traffic.
Solution Approach 2:
The patent uses the density ratio as a feedback mechanism to guide the formation of virtual groups. By continuously evaluating the ratio of traffic leaving a virtual group to the number of VNFs within it, the system can iteratively refine group compositions to achieve optimal balance between traffic maximization and cluster integrity.
3Productivity
If VNFs are separated into different software clusters, then more VNFs can be provisioned on fewer computing devices, but communication latency increases and network bandwidth is consumed for VNFs that should communicate
Solution Approach 1:
The patent merges VNFs into densely-connected virtual groups that are then provisioned together on the same physical group (server or rack). This merging ensures that VNFs requiring frequent communication remain co-located physically, reducing communication latency while still achieving resource consolidation across the infrastructure.
Solution Approach 2:
By changing the optimization criterion to density-based grouping, the patent ensures that VNFs with high intercommunication requirements are kept together in the same virtual group, which is then provisioned on the same physical group. This parameter change directly addresses the latency issue by prioritizing communication-intensive VNF relationships.
4Productivity
If conventional clustering techniques are used, then hardware clusters can be efficiently utilized, but the system does not account for power efficiency and bandwidth conservation in the allocation decisions
Solution Approach 1:
The patent merges VNFs into virtual groups based on density criteria and provisions entire virtual groups on the same physical group (server or rack). This merging reduces the total number of physical devices required, thereby reducing overall power consumption while maintaining efficient hardware utilization within the deployed devices.
Solution Approach 2:
The patent introduces density ratio as the optimization parameter, which indirectly optimizes power efficiency by creating more compact virtual groups. This parameter change leads to better resource consolidation, fewer active physical devices, and reduced power consumption compared to conventional bandwidth-based approaches.
Data Source
AI summary
A virtualized network function (VNF) provisioning system includes a plurality of computing device that are provided in a physical topology and that are coupled to a network. A management device is coupled to the computing device through the network. The management device receives VNF system information for a VNF system through the network. The VNF system includes a plurality of VNFs that are provided in a virtual topology. The management device then determines a plurality of physical groups from the physical topology that each identify at least one computing device. The management device then determines a plurality of virtual groups from the virtual topology based on a ratio of traffic that will leave each virtual group and a number of VNFs in each virtual group. The management device then provides the VNFs in a first virtual group on the at least one computing device in a first physical group.