Flooding Topology Computation for Network Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As networks grow, interior gateway protocol (IGP) routing convergence becomes slower due to synchronization issues in link-state databases during link failures, leading to delayed network convergence and increased bandwidth consumption.
Innovation Solution
A method for computing a flooding topology (FT) in network devices, involving selecting a node, initializing the FT, and iteratively adding nodes and links to ensure all nodes are connected with minimized degrees and links, allowing for simultaneous deterministic computation across multiple nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional IGP flooding topology is used to ensure all nodes receive routing information, then network coverage is complete, but flooding traffic increases and convergence time increases
Solution Approach 1:
The patent extracts only the essential nodes and links needed for routing information propagation by computing a minimum spanning tree of the flooding topology. This removes redundant links and nodes from the flooding process, reducing bandwidth consumption while maintaining complete routing information delivery to all nodes in the network.
Solution Approach 2:
The patent creates a simplified copy of the network topology in the form of a minimum spanning tree structure. This copied topology contains only the necessary paths for flooding, allowing routing information to be propagated efficiently without transmitting over all possible links in the original topology, thus reducing bandwidth usage while preserving delivery reliability.
2Reliability
If traditional IGP flooding is used to propagate routing information, then all nodes are covered, but convergence time increases
Solution Approach 1:
The patent extracts the critical path structure by computing a minimum spanning tree, removing redundant transmission paths. This extraction reduces the number of flooding steps required for routing information to reach all nodes, directly reducing convergence time while maintaining complete coverage through the optimized topology structure.
Solution Approach 2:
The patent dynamically adapts the flooding topology by computing a minimum spanning tree based on current network conditions and node degrees. This dynamic optimization allows the flooding process to follow the most efficient paths adaptively, reducing convergence time compared to static flooding over the complete topology.
3Loss of energy
If minimum spanning tree optimization is applied to reduce flooding traffic, then bandwidth consumption decreases, but device complexity increases
Solution Approach 1:
The patent implements a universal minimum spanning tree computation algorithm that can be applied to any network topology. This multi-functional approach handles various network configurations (different node degrees, link configurations) using the same algorithmic framework, making the complexity management systematic and reusable across different network scenarios rather than requiring custom solutions for each case.
Solution Approach 2:
The patent enables network devices to autonomously compute their own flooding topology using the minimum spanning tree algorithm based on locally available topology information. This self-service capability reduces the need for centralized control and complex coordination, allowing devices to independently optimize their flooding behavior while maintaining overall network efficiency.
Data Source
AI summary
Methods of computing a flooding topology (FT) for a network are presented. The methods include a process for computing a FT that includes all nodes in the network and a process for ensuring that all nodes in the FT have at least two links in the FT. Some of the methods minimize a number of links of the nodes in the FT. Some of the methods also constrain some of the nodes in the FT to a maximum number of links. Some of the methods compute a first FT for nodes whose maximum number of links in the FT equal their number of links in the network, then compute a second FT for remaining nodes in the network, then combines the two FTs to compute a complete FT for the network.