Multi-Threaded SDN Simulation for Lightweight Node Benchmarking

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Implementing an entire software-defined network (SDN) or its relevant portion for performance benchmarking is expensive and resource-intensive.

Innovation Solution

A lightweight method involving a simulator instance on a second SDN node to generate parallel threads simulating connections with other nodes, transmitting service messages to benchmark the performance of a first SDN node.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If an entire SDN or relevant portion is implemented for performance benchmarking, then measurement precision is improved, but device complexity and cost increase

Engineering Contradiction:
Improveperformance benchmarking accuracyVSAvoidSDN implementation complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a simulated copy of the SDN environment using a lightweight simulator that replicates essential SDN node behaviors and message exchanges. Instead of deploying actual SDN infrastructure, the simulator generates synthetic service messages that mimic real SDN traffic patterns, enabling accurate performance benchmarking without the complexity of a full SDN deployment.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent employs disposable, lightweight simulated SDN nodes that can be rapidly created and destroyed for benchmarking purposes. These simulated nodes are inexpensive computational objects that generate and process service messages temporarily during testing, eliminating the need for persistent, expensive physical SDN infrastructure while maintaining benchmarking validity.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

2Reliability

If an entire SDN is implemented for performance benchmarking, then reliability of benchmarking results is improved, but loss of energy and resources increases

Engineering Contradiction:
Improvebenchmarking result reliabilityVSAvoidcomputational resource consumption
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts only the essential elements needed for performance benchmarking from a complete SDN system. The simulator isolates and replicates only the critical message exchange patterns between SDN nodes, excluding unnecessary infrastructure components. This extraction maintains benchmarking reliability by focusing on core functional interactions while dramatically reducing computational resource consumption.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements partial action by generating a sufficient number of simulated service messages to achieve statistically reliable benchmarking results without creating an exhaustive replica of real-world SDN traffic volumes. The simulator produces just enough synthetic traffic to accurately measure node performance characteristics, avoiding the excessive resource consumption that would result from implementing complete SDN infrastructure.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4614921A1Multi-threaded simulator for performance benchmarking in software-defined networks
Publication Date: 2025.09.10 NOKIA SOLUTIONS & NETWORKS OY
  • EP4614921A1 patent drawingFigure 1
  • EP4614921A1 patent drawingFigure 2
  • EP4614921A1 patent drawingFigure 3

AI summary

Performance of a node-under-test in a software-defined network (SDN) is benchmarked by implementing a simulator instance in another SDN node that transmits series of messages over multiple, parallel threads to the node-under-test. When the node-under-test is an SDN network management block (NMB), the other SDN node may be an SDN controller (SDNC), where each thread simulates a real-world connection between the NMB and a different real-world SDNC. When the node-under-test is an SDNC, the other SDN node may be an SDN network service gateway (NSG), where each thread simulates a real-world connection between the SDNC and a different real-world NSG. When the node-under-test is an NSG, the other SDN node may be user equipment, where each thread simulates a real-world connection between the NSG and different real-world user equipment. The disclosure enables lightweight, explicit performance benchmarking of SDN nodes without having to implement an entire SDN in the testing environment.