Neural Network Traffic Decomposition for 5G Slicing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting future per-service traffic consumption in 5G networks are resource-intensive and time-consuming, making accurate predictions impractical without excessive computational and time resources.
Innovation Solution
A distributed network traffic data decomposition method using a neural network to analyze and convert aggregate network traffic data into a regular grid pattern, employing 3D deformable and 2D convolutions to extract spatiotemporal correlations, and predicting future per-service traffic consumption without deep packet inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep packet inspection is used to estimate service level demands, then measurement precision is improved, but use of energy and loss of time increase significantly
Solution Approach 1:
The patent extracts only the necessary features from aggregate traffic data that are sufficient for service level demand estimation, avoiding the need for comprehensive deep packet inspection. By selecting and analyzing only relevant data elements, the system achieves adequate measurement precision with significantly reduced computational energy consumption.
Solution Approach 2:
The patent applies partial action by using simplified analysis methods that process only a portion of the available data with sufficient depth for estimation purposes. Rather than performing complete deep packet inspection on all traffic data, the system applies targeted analysis to aggregate metrics, achieving acceptable precision with reduced resource expenditure.
2Measurement precision
If deep packet inspection is used to estimate service level demands, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent extracts only the essential components needed for service level demand estimation from the traffic data, avoiding time-consuming comprehensive inspection. By focusing on key aggregate metrics rather than analyzing every packet, the system achieves sufficient measurement precision while dramatically reducing processing time.
Solution Approach 2:
The patent implements partial action by applying simplified estimation techniques to aggregate data that capture the essential patterns for service level demands. This partial analysis approach provides adequate precision for network management decisions without the excessive time investment required for complete deep packet inspection.
3Productivity
If simplified methods are used for traffic analysis, then use of energy and loss of time are reduced, but measurement precision deteriorates
Solution Approach 1:
The patent changes the parameters of analysis by working with aggregate traffic metrics and derived features rather than raw packet-level data. By transforming the data into appropriate aggregate representations and applying neural network analysis on these transformed parameters, the system achieves high prediction accuracy with significantly improved productivity and updated frequency.
Solution Approach 2:
The patent transitions from analyzing individual packet dimensions to analyzing aggregate traffic patterns across multiple dimensions simultaneously. By using neural networks to process multi-dimensional aggregate metrics (traffic volume, patterns, correlations across multiple sources), the system achieves accurate predictions while maintaining high processing efficiency and update frequency.
Data Source
AI summary
To be able to adequately provide desired services over a 5G mobile service network, the 5G communication infrastructures requires a much-improved flexibility in resource management. Network operators are foreseen to deploy network slicing, by isolating dedicated resources and providing customised logical instances of the physical infrastructure to each service. A critical operation in performing management and orchestration of network resources is the anticipatory provisioning of isolated capacity to each network slice. Accordingly, it is necessary to obtain an estimate of service level demands. However, the estimation of such service level demands is typically obtained via deep packet inspection, which is a resource intensive and time-consuming process. Therefore, it is typically not possible to provide updated accurate estimates at a frequency suitable for use in accurate prediction of a future per-service traffic consumption, without an undesirable level of computational and time resources being required. The present invention provides a distributed network traffic data decomposition method which makes use of a neural network to provide an accurate future per-service traffic consumption prediction without deep-packet inspection or another resource intensive analysis method.