Network Probe Placement Optimization in Leaf-Spine Cloud Fabric

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In network probe placement, existing technologies face challenges in minimizing the number of probes and probe traffic while maintaining comprehensive network coverage, especially in dense cloud fabrics like 5G and 6G edge compute platforms, which can lead to undesirable network performance and increased costs.

Innovation Solution

The implementation of a system that determines the optimal placement and distribution of Two-Way Active Measurement Protocol (TWAMP) probes using a directed acyclic graph and Monte Carlo sampling to minimize the number of probes required for full network coverage, ensuring statistical certainty and reducing bandwidth consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If active probes are deployed to cover every physical network path, then network coverage is improved, but network performance deteriorates due to excessive probe traffic

Engineering Contradiction:

Improvenetwork coverageVSAvoidnetwork performance

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent applies partial action by deploying probes only at strategically selected network nodes rather than every node. The system determines optimal probe placement locations that provide sufficient network coverage while avoiding the performance degradation caused by excessive probe deployment at all nodes.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent changes the parameter of probe deployment from universal (every node) to selective (optimal nodes). By using graph theory algorithms to analyze network topology and identify critical nodes, the system transforms the probe deployment strategy to achieve coverage with fewer probes, thereby reducing probe traffic volume and maintaining network performance.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If the number of probes is reduced to improve network performance, then bandwidth consumption is improved, but network coverage measurement precision deteriorates

Engineering Contradiction:

Improvebandwidth consumptionVSAvoidnetwork coverage measurement

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent changes the placement location parameter of probes using graph theory optimization. By identifying nodes with maximum topological importance (high betweenness centrality, connectivity), the system ensures that fewer probes can still capture representative network paths and provide accurate coverage measurements.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces brute-force probe deployment (mechanical approach of placing probes everywhere) with an algorithmic approach using graph theory and Monte Carlo simulations. This substitution allows the system to mathematically determine optimal probe locations that maintain measurement precision while reducing probe count.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If probes are placed at every network node to ensure full coverage, then network coverage is improved, but device complexity and cost increase

Engineering Contradiction:

Improvenetwork coverageVSAvoidprobe distribution complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the deployment strategy from comprehensive (all nodes) to optimized (critical nodes only). By using graph theory to identify and prioritize network nodes based on their topological importance, the system reduces the number of probes needed while maintaining coverage effectiveness, thereby simplifying device distribution and reducing costs.

Inventive Principle:

Principle #35Parameter changes

4Reliability

If routing algorithms are used to direct probe traffic, then probe delivery is improved, but probe traffic may not follow the same path as customer traffic, reducing measurement accuracy

Engineering Contradiction:

Improveprobe deliveryVSAvoidcustomer experience measurement

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The patent inverts the approach by not trying to force probe traffic to follow customer traffic paths through routing manipulation. Instead, it selects probe source and destination nodes such that probe paths naturally represent customer traffic patterns, using graph theory to identify pairs of nodes that are likely endpoints of customer flows.

Inventive Principle:

Principle #13The other way round (Inversion)

Data Source

PatentUS11533246B1Network probe placement optimization

Publication Date: 2022.12.20 AT&T INTELLECTUAL PROPERTY I L P

AI summary

Intelligent network probe placement optimization (e.g., using a computerized tool) is enabled. A method can comprise determining, from an inventory database, physical interfaces and service paths for data traffic to be monitored, according to a two-way active management protocol, with respect to network interfaces between cloud compute elements in a leaf-spine cloud fabric and radio access network equipment, based on the cloud compute elements, determining directed acyclic graph information representative of a directed acyclic graph of connections between the cloud compute elements and other cloud compute elements in the leaf-spine cloud fabric other than the cloud compute elements, and based on the directed acyclic graph information, determining a number and a distribution of probes to be employed at at least some of the physical interfaces and the service paths to monitor a parameter of the data traffic according to the two-way active management protocol.