Network Service Scheduling Using Traffic Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current bandwidth scheduling in content delivery networks (CDNs) is inaccurate due to allocation based on service priority, leading to surplus bandwidth for high-priority services and insufficient bandwidth for low-priority services, resulting in improper bandwidth usage.
Innovation Solution
A network service scheduling method and apparatus that predicts future network traffic using historical data and a machine learning model, allowing for precise scheduling of network resources based on predicted traffic patterns, thereby optimizing bandwidth allocation across services.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If bandwidth is allocated according to service priority, then high-priority services receive sufficient bandwidth, but low-priority services receive insufficient bandwidth and high-priority services may have surplus bandwidth
Solution Approach 1:
The system performs preliminary actions by predicting future network traffic before the actual traffic occurs. Historical traffic data is analyzed to forecast upcoming traffic patterns, allowing the bandwidth allocation system to prepare and adjust bandwidth distribution in advance, thereby avoiding both surplus and insufficient bandwidth allocation
Solution Approach 2:
The bandwidth allocation system transitions from a static priority-based approach to a dynamic approach that continuously adapts to changing traffic patterns. The system uses real-time traffic monitoring and prediction to dynamically adjust bandwidth allocation, ensuring that bandwidth is allocated according to actual needs rather than fixed priority levels
2Reliability
If bandwidth is allocated according to service priority, then high-priority services are guaranteed bandwidth, but the overall bandwidth scheduling accuracy deteriorates
Solution Approach 1:
The system performs preliminary traffic prediction using historical data and machine learning models to forecast future traffic patterns. This preliminary action enables the system to determine accurate bandwidth requirements before actual traffic occurs, thereby improving scheduling accuracy while maintaining reliability for high-priority services
Solution Approach 2:
The system implements a feedback mechanism that continuously monitors actual traffic patterns and compares them with predicted patterns. This feedback is used to refine and update the prediction models, improving the accuracy of bandwidth scheduling over time while ensuring that high-priority services continue to receive guaranteed bandwidth
3Reliability
If bandwidth is allocated according to service priority, then high-priority services receive sufficient bandwidth, but bandwidth utilization efficiency deteriorates
Solution Approach 1:
The system transitions from static priority-based bandwidth allocation to dynamic allocation that adapts to real-time and predicted traffic patterns. This allows the system to optimize bandwidth utilization efficiency by allocating bandwidth according to actual needs while maintaining the reliability guarantees for high-priority services through the prediction-based approach
Data Source
AI summary
Aspects of the disclosure provide a method and an apparatus for network service scheduling. The apparatus includes interface circuitry and processing circuitry. The interface circuitry receives network traffic data from devices in a content delivery network (CDN) that provides network services. The processing circuitry obtains historical network traffic data of the network services. The historical network traffic data includes network traffic measures of the network services in past time units from a present moment. The processing circuitry predicts future network traffic of the network services in a next time unit after the present moment according to the historical network traffic data of the network services. Then the processing circuitry schedules network resources for the network services according to the predicted future network traffic.