Cross-Layer Controller for SDN Flow Entry Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing layered management mode in SDN networks is inefficient in acquiring service paths due to complex processes and low efficiency, particularly in managing IP and optical layer resources separately.
Innovation Solution
A cross-layer information model is introduced where both IP layer and optical layer resources are considered at the same layer, allowing for simultaneous management and reducing the time complexity of service path computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a layered management mode is used with separate IP layer controller and optical layer controller, then resource management is organized in layers, but the service path acquisition process becomes very complex and efficiency is very low
Solution Approach 1:
The patent merges the IP layer controller and optical layer controller into a unified controller that manages both layers simultaneously. This consolidation eliminates the complex interaction and multiple acquisition steps between separate layered controllers, directly resolving the contradiction by combining control functions to improve service path acquisition efficiency while maintaining organized resource management.
2Adaptability or versatility
If IP layer controller and optical layer controller operate independently in layered mode, then each layer manages its own resources, but the overall service path computation time complexity is high
Solution Approach 1:
The patent combines IP layer and optical layer control functions into a single unified controller that computes service paths across both layers simultaneously. This merging allows the system to maintain the adaptability of layered resource management while reducing computation time complexity from O(n*m) to O(n+m) by performing unified path computation rather than sequential layered acquisition.
3Device complexity
If separate layered controllers are used for IP and optical layers, then resource management is specialized per layer, but the process of acquiring service path is very complex
Solution Approach 1:
The patent merges separate IP layer and optical layer controllers into a unified controller that handles both layers. This consolidation simplifies the service path acquisition process from a complex multi-step interaction between separate controllers to a single unified operation, making the process easier to operate while maintaining the specialized resource management capabilities of each layer.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This application discloses a flow entry generating method and apparatus. The method includes: receiving a service path establishment request, where the service path establishment request includes a constraint, a source device, and a target device; finding, in a data transmission network according to a cross-layer information model, a service path that meets the constraint and is from the source device to the target device, where the cross-layer information model is a model that describes, at a same layer, an overall topology relationship between an IP layer and an optical layer; and generating a respective corresponding flow entry for each forwarding device in the service path, and sending the flow entry to a corresponding forwarding device. In the foregoing solutions, according to a cross-layer information model in which an IP layer resource and an optical layer resource are located at a same layer, both the IP layer resource and the optical layer resource are considered during service path computation, which can greatly reduce time complexity of computing the service path, and improve efficiency of computing the service path, thereby improving efficiency of generating a flow entry.