Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An intelligent energy-saving control method based on data center network traffic prediction and learning

A data center network, energy-saving control technology, applied in data exchange networks, neural learning methods, biological neural network models, etc., can solve problems such as efficiency and optimization results cannot be effectively guaranteed, and achieve multi-material network flow problems. Accurately predict and optimize the effect of bandwidth allocation

Active Publication Date: 2021-10-12
EAST CHINA NORMAL UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

To solve this problem, existing solutions often use heuristic strategies to save energy as much as possible in the routing process, but their efficiency and optimization results cannot be effectively guaranteed

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • An intelligent energy-saving control method based on data center network traffic prediction and learning
  • An intelligent energy-saving control method based on data center network traffic prediction and learning
  • An intelligent energy-saving control method based on data center network traffic prediction and learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] The present invention proposes a hybrid superimposed LSTM model for demand forecasting, the input part of the model is composed of two superimposed LSTM layers. In order to match the dimension of the label, a series of fully connected layers are configured in the output part of the model to adjust the dimension of the model output. LSTM models overcome the vanishing gradient problem by introducing cell states and control gates in each cell.

[0024] See attached figure 1 , the present invention designs a set of deep reinforcement learning methods based on Deep Deterministic Policy Gradient (Deterministic Policy Gradient) algorithm for the multi-object network flow problem, including a traffic prediction module (RNN) and a traffic traffic optimization module (RL) with the current The data center network traffic prediction and learning system of network, network topology, topology and routing architecture, its deep reinforcement learning and network optimization specific...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an intelligent energy-saving control method based on data center network traffic prediction and learning, which is characterized in that a mixed and superimposed neural network model is used to predict the network traffic in the data center network, and deep reinforcement learning of the DDPG algorithm is used to optimize the network The bandwidth allocation and routing selection in the network realize the energy-saving control of the data center network and the accurate prediction of future network traffic. Compared with the prior art, the present invention has accurate prediction results and intelligent energy-saving control, greatly optimizes bandwidth allocation and routing selection in the network, and realizes effective and accurate prediction of future network traffic. A good solution to the multi-object network flow problem in the data center network.

Description

technical field [0001] The invention relates to the technical field of network optimization and flow prediction, in particular to an intelligent energy-saving control method based on data center network flow prediction and learning. Background technique [0002] With the widespread application of cloud computing in search engines, social media, e-commerce, etc., in recent years, Data Center Networking (DCN) has become an important network structure, which provides large-scale storage and high-performance Compute provides warehouse-level computing services. The widespread use of cloud computing and data center networks has also brought about a synchronous increase in energy consumption. The industry can no longer ignore its energy consumption. Some bandwidth allocation and routing optimization methods for energy saving in data center networks have been proposed one after another. . Most of the existing bandwidth allocation and routing optimization methods are offline analys...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): H04L12/24G06N3/04G06N3/08
CPCG06N3/049G06N3/08H04L41/0896H04L41/14H04L41/147
Inventor 汪洋王廷厉宇桐
Owner EAST CHINA NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products