Network Slice Deployment via Deep Learning Congestion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques fail to guarantee network slices for critical infrastructure users during network congestion, leading to poor user experience and lost network traffic due to resource consumption.
Innovation Solution
A deep learning model is used to detect network congestion and automatically deploy a new network slice, selecting the region, duration, and type based on media and weather data to ensure resource allocation for critical infrastructure users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network slices are deployed to ensure service for critical infrastructure users, then service reliability is improved, but network resources are consumed that could be used for other purposes during congestion
Solution Approach 1:
The system performs preliminary actions by predicting network congestion using deep learning models before actual congestion occurs. The AI model analyzes historical data, traffic patterns, and network conditions to forecast congestion events, allowing the system to proactively deploy network slices in advance rather than reacting after congestion has occurred, thus optimizing resource allocation.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring network conditions, user experience metrics, and congestion levels. This feedback is fed back into the deep learning model to refine predictions and adjust network slice deployment decisions in real-time, ensuring optimal resource allocation that balances service reliability with resource conservation.
2Ease of operation
If network slices are automatically deployed during congestion, then user experience is improved, but system complexity increases due to automated decision-making
Solution Approach 1:
The system performs self-service by automatically detecting network congestion and deploying appropriate network slices without requiring manual intervention. The deep learning model autonomously analyzes network conditions, makes predictions about congestion, and triggers slice deployment based on predefined policies, eliminating the need for operator involvement in real-time decision-making.
Solution Approach 2:
The system replaces manual mechanical decision-making with automated AI-based systems. Instead of network operators manually monitoring and deciding when to deploy network slices, a deep learning model automatically processes network data, predicts congestion scenarios, and executes deployment actions, substituting human operational complexity with automated intelligent systems.
3Productivity
If manual network slice deployment is used, then resource allocation can be optimized, but response time during congestion increases
Solution Approach 1:
The system performs preliminary actions by predicting network congestion before it occurs and proactively deploying network slices in advance. This eliminates the need to wait for actual congestion to manifest, significantly reducing response time while maintaining optimized resource allocation through AI-driven predictions.
Solution Approach 2:
The system replaces slow manual resource allocation processes with automated AI-based decision-making that operates in real-time. The deep learning model processes network data and executes deployment decisions much faster than manual operations, dramatically reducing response time while maintaining optimization through intelligent algorithms.
Data Source
AI summary
A device may receive network congestion data associated with a network providing a network slice to a user device and that includes data identifying a number of users, an average throughput, a latency, and one or more key performance indicators associated with the network. The device may receive media data associated with the user device and may receive weather data associated with the user device. The device may process the number of users, the average throughput, the latency, and the one or more key performance indicators, with a deep learning model, to generate a congestion decision for the network. The device may selectively activate a new network slice for the user device based on the congestion decision indicating congestion in the network, or may maintain the network slice for the user device based on the congestion decision indicating no congestion in the network.