Virtual CDN Scale Adjustment via ARIMA Flow Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional CDN networks face high energy consumption due to redundant hardware design and inefficiencies in predicting network flow, leading to suboptimal server scaling and increased carbon emissions.
Innovation Solution
An energy-saving deployment method for virtual CDNs that uses an ARIMA prediction model to dynamically adjust server scale based on historical flow data, incorporating a differential order and parameters to forecast peak flow, and adds redundant servers to manage peak loads, optimizing server utilization and energy efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant hardware servers are deployed to meet peak network flow, then service reliability is improved, but energy consumption increases
Solution Approach 1:
The patent applies dynamics by transitioning from static hardware servers to dynamic virtual servers that can be scaled up or down based on real-time network flow conditions. The system dynamically adjusts the number of active servers from 10,000 during peak flow to a smaller number during non-peak periods, maintaining service reliability when needed while reducing energy consumption when demand is lower.
Solution Approach 2:
The patent changes the parameter of server capacity from fixed hardware-based capacity to variable virtual-based capacity. By using virtualization technology, the system can change the number of active servers as a parameter based on predicted network flow, allowing the system to operate with fewer servers during non-peak times while maintaining the ability to scale up when needed.
2Measurement precision
If BP neural network is used to predict network flow, then prediction capability is improved, but computational complexity and convergence time increase
Solution Approach 1:
The patent uses a simpler, more efficient prediction model (ARIMA) that is computationally lighter and faster to converge compared to complex BP neural networks. While BP neural networks offer sophisticated prediction capabilities, the patent chooses a more practical approach that provides sufficient prediction accuracy without the computational burden and convergence issues of neural networks.
Solution Approach 2:
The patent changes the prediction model from complex neural network parameters to simpler ARIMA model parameters (p, d, q values). This parameter change results in a model that is easier to compute, requires less training data, and converges faster while still providing effective network flow predictions for CDN scaling decisions.
3Device complexity
If average flow data is used to determine CDN scale, then system simplicity is maintained, but peak flow handling capability deteriorates
Solution Approach 1:
The patent applies preliminary action by using predicted peak flow data to proactively determine CDN scale before actual peak flow occurs. Instead of reacting to average flow data, the system uses prediction models to forecast future peak conditions and scales the CDN accordingly in advance, ensuring adequate capacity is available when needed without excessive complexity in the decision-making process.
Data Source
AI summary
The present invention discloses an energy-saving deployment method of a virtual CDN. According to the historical flow data of the virtual CDN and the prediction model (ARIMA) in the controller, the network peak flow in the next time period is predicted. Next, the scale of the virtual CDN system at the next moment is calculated according to the peak flow. Meanwhile, several redundant servers are added to correct the prediction error. The network flow is aggregated to the desired virtual servers based on the calculation of the controller through a load balancer. In this way, the utilization rate of the virtual CDN system can be increased, and the energy consumed due to the higher utilization rate of the CDN system is saved.