Dynamic Network Topology Optimization for Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current physical data communications topologies between computing nodes often lead to network congestion and reduced performance due to uneven network traffic distribution, where some links are overloaded while others are underutilized, resulting in increased latency.
Innovation Solution
A method that detects network packets, assigns packet weights based on attributes, determines node pair traffic weights, and reconfigures physical links between nodes to optimize traffic distribution by allocating additional links where needed and removing unnecessary ones, using a topology optimization ASIC and management module.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a balanced physical data communications topology is used to connect computing nodes, then the network structure is simple and easy to manage, but network latency increases and performance decreases due to uneven traffic distribution causing link congestion
Solution Approach 1:
The patent implements dynamic topology reconfiguration by continuously monitoring packet weights and traffic patterns between node pairs, then adjusting the number of physical links in real-time. This transforms the static balanced topology into a dynamic adaptive structure that responds to changing traffic demands, resolving the contradiction between structural simplicity and performance efficiency
Solution Approach 2:
The system changes the parameter of link quantity between node pairs based on measured traffic weights. By adjusting the number of physical links from a uniform balanced configuration to a variable configuration where high-traffic pairs have more links and low-traffic pairs have fewer links, the system optimizes throughput while maintaining manageable complexity through automated control
2Productivity
If additional physical links are added between high-traffic node pairs to reduce congestion, then network performance improves, but device complexity and configuration difficulty increase
Solution Approach 1:
The topology optimization system performs self-configuration by automatically detecting traffic patterns, calculating optimal link distributions, and reconfiguring the physical topology without manual intervention. This self-service capability resolves the contradiction by enabling complex adaptive topologies to be maintained through automated feedback loops rather than manual configuration
Solution Approach 2:
The system establishes a closed-loop feedback mechanism where packet weights are continuously measured, link configurations are adjusted based on these measurements, and the results are monitored to further refine the topology. This feedback-driven approach enables the system to automatically optimize throughput while managing complexity through iterative improvement rather than complex static design
Data Source
AI summary
Methods, apparatus, and products are disclosed for optimizing a physical data communications topology between a plurality of computing nodes, the physical data communications topology including physical links configured to connect the plurality of nodes for data communications, that include carrying out repeatedly at a predetermined pace: detecting network packets transmitted through the links between each pair of nodes in the physical data communications topology, each network packet characterized by one or more packet attributes; assigning, to each network packet, a packet weight in dependence upon the packet attributes for that network packet; determining, for each pair of nodes in the physical data communications topology, a node pair traffic weight in dependence upon the packet weights assigned to the network packets transferred between that pair of nodes; and reconfiguring the physical links between each pair of nodes in dependence upon the node pair traffic weights.