Neural Network Learning for Identifying Bad Network Ports
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Monitoring the performance of individual network ports in large communication networks is economically unfeasible due to the resource-intensive use of physical sensors, making it difficult to identify and address ports responsible for packet loss or delay.
Innovation Solution
A neural network learning method that models network paths and iteratively analyzes existing data to identify bad ports by determining weight factors using a binary value function, allowing for the identification of ports causing packet loss or delay without the need for physical sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical sensors are deployed to monitor each network port, then measurement precision of port performance is improved, but device complexity and resource consumption increase significantly
Solution Approach 1:
The patent creates a virtual copy of the network topology and uses neural networks to replicate the function of physical sensors. Instead of deploying actual sensors to each port, the system uses software-based monitoring agents that collect performance data and feed it to neural network models, which then identify problematic ports through pattern recognition.
Solution Approach 2:
The patent replaces the mechanical/physical sensor system with an information-processing system based on neural networks. The physical sensors that would directly measure port performance are substituted with software agents that collect data and neural network algorithms that analyze patterns to identify bad ports, eliminating the need for physical hardware at each port.
2Measurement precision
If physical sensors are deployed to monitor each network port, then measurement precision of port performance is improved, but resource consumption increases significantly
Solution Approach 1:
The patent creates a virtual monitoring system that copies the essential function of physical sensors through software. Neural network models process collected data to identify problematic ports, eliminating the need for energy-intensive physical sensor hardware at each network port while maintaining measurement capability.
Solution Approach 2:
The patent uses lightweight software-based monitoring agents instead of expensive, energy-consuming physical sensors. These software agents consume minimal computational resources and can be deployed across numerous ports without the hardware overhead, making the monitoring system economically feasible for large-scale networks.
3Productivity
If traditional monitoring methods are used to identify bad ports, then device complexity is reduced, but productivity in identifying and addressing packet loss issues decreases
Solution Approach 1:
The patent implements preliminary action by continuously training and updating neural network models with historical network performance data. The system learns patterns of packet loss and port degradation in advance, so when issues occur, the pre-trained models can quickly identify problematic ports without requiring complex real-time analysis, thus improving productivity while managing complexity through offline preparation.
Solution Approach 2:
The patent incorporates feedback mechanisms where the neural network system continuously receives performance data from network ports, processes it through learning algorithms, and generates identification results that are fed back to network operators. This closed-loop system improves productivity by automatically learning from past performance patterns and refining its identification accuracy over time without requiring manual intervention or complex manual analysis procedures.
Data Source
AI summary
A computational method and system for identifying bad ports in a network may use a neural network learning function based on available network path data that is already collected. In this manner, bad ports in the network may be identified without having to measure each individual port using sensors.