Online network flow prediction method and system based on newly increased flow number characteristics
A network traffic and number technology, applied in the field of online real-time network traffic prediction by neural network, can solve the problems of inability to process the data characteristics of traffic data flow, and the regression prediction model cannot achieve good traffic time series prediction effect, and achieve accurate time point and training. good round effect
Active Publication Date: 2022-01-28
COMP NETWORK INFORMATION CENT CHINESE ACADEMY OF SCI
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0007]
The purpose of the present invention is to solve the problem that the existing regression prediction model cannot achieve a good flow time series prediction effect, and solve the problem that the existing flow prediction system cannot process flow data in real time online and obtain flow data characteristics
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0025] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the above description. Although the present embodiment shows exemplary embodiments of the present disclosure, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.
[0026] figure 1 It is a schematic flow chart of an online traffic prediction method based on the newly added flow number feature provided by the embodiment of the present invention. Such as figure 1 As shown, the embodiment of the present invention mainly includes two parts, a prediction model part based on the new flow number feature and a real-time streaming data processing part of online network traffic data.
[0027] For th...
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
Login to View More Abstract
The invention relates to an online network flow prediction method and system based on newly increased flow number characteristics, and the method comprises the steps: collecting and analyzing real flow data of a network, transmitting the real flow data to a message queue, enabling the real flow data to flow into a flow processing engine, and carrying out the online calculation to obtain a real-time flow time sequence and a newly increased flow number characteristic sequence provided by the invention, i.e., the number of newly added network flows in each time granularity; and simultaneously putting the two sequences into a network traffic prediction model with multi-feature input and multi-time step output for traffic prediction. According to the method, the real-time performance of traffic prediction can be guaranteed, that is, the time length of the predicted traffic sequence is greater than the sum of the acquisition time, the stream processing time and the prediction time, and the time length can be fed back to the network traffic controller in time, so the network controller can regulate and control the network traffic and make a load balancing decision; the average error of the obtained prediction result is reduced by more than 20% compared with that of the LSTM model which only inputs the flow sequence, and the sudden rising, falling and peak conditions of the real flow can be better predicted.
Description
technical field [0001] The present invention relates to the fields of network traffic forecasting, deep learning, deep message detection, and data processing. Specifically, the present invention relates to online real-time prediction of the network through a neural network after obtaining the characteristics of the traffic and the number of new flows based on the processing data of the deep message detection. Method and system for flow. Background technique [0002] After decades of development, the Internet has formed a global communication network infrastructure. With the access of a large number of network devices in the Internet and the further improvement of users' requirements for network service quality, network traffic has increased exponentially, and emergencies have also increased. In order to ensure the quality of service and optimize the deployment and allocation of network resources, fast, real-time and high-precision prediction of network traffic online is esp...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): H04L43/0876H04L41/14H04L41/147G06N3/04G06N3/08
CPCH04L43/0876H04L41/145H04L41/147G06N3/08G06N3/048G06N3/044Y02D30/50
Inventor 李少鹤周旭宋俊平徐陆阳李菁菁
Owner COMP NETWORK INFORMATION CENT CHINESE ACADEMY OF SCI