Metropolitan Optical Network Traffic Prediction Using Node Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic prediction methods for metropolitan optical networks are inaccurate due to static resource allocation and failure to consider tidal phenomena, which cause different traffic patterns among nodes in various regions.
Innovation Solution
A traffic prediction method that classifies nodes in a metropolitan optical network into multiple node sets based on location, using deep learning to establish a traffic prediction model for each node set that accounts for tidal phenomena.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If nodes are predicted as a whole without classification, then the prediction process is simple, but the prediction accuracy is low due to ignoring tidal phenomena in different regions
Solution Approach 1:
The patent segments the metropolitan optical network nodes into multiple node sets based on their locations and tidal phenomena characteristics. Each node set is predicted using a dedicated traffic prediction model trained on historical data specific to that region's tidal patterns, rather than using a single unified model for all nodes. This segmentation approach improves prediction accuracy by capturing regional differences while maintaining manageable model complexity through modular architecture.
2Adaptability or versatility
If static network resource allocation is used, then the system is simple to manage, but service congestions occur due to inability to adapt to varying network loads
Solution Approach 1:
The patent implements dynamic network resource allocation that adapts to varying network loads by utilizing the traffic prediction results from the deep learning models. The system dynamically adjusts resource allocation based on predicted traffic patterns and tidal phenomena, enabling the network to proactively respond to load changes rather than using static allocation. This dynamic approach prevents service congestions while maintaining system manageability through automated prediction-driven control.
3Measurement precision
If deep learning with historical time-series data is used for each node set, then prediction accuracy improves, but computational requirements and data processing complexity increase
Solution Approach 1:
The patent divides the large-scale data processing task into smaller, manageable segments by creating multiple node sets based on geographic locations and tidal phenomena. Each node set has its own dedicated deep learning model trained on localized historical time-series data. This segmentation reduces the computational burden on individual models compared to training a single model on all network data, while still achieving high prediction accuracy through region-specific training data that captures local traffic patterns.
Data Source
AI summary
Disclosed are a traffic prediction method for a metropolitan optical network and related devices. The traffic prediction method may include: classifying nodes in the metropolitan optical network into multiple node sets based on locations of the nodes; for nodes in each node set, inputting temporal traffic data of the nodes into a traffic prediction model corresponding to the node set to obtain traffic prediction results of the nodes in the node set. In the method, the traffic prediction model is obtained by deep learning using historical time-series traffic data of the node sets as a training set.