A sketch-based network measurement method based on reinforcement learning

Abstract

The invention relates to a network measurement method of Sketch based on reinforcement learning. This method uses the reinforcement learning method to enable Sketch to find the optimal resource and measurement accuracy balance in the real network environment, so as to adapt to the changing network environment, improve the measurement performance, and reduce the configuration burden of the administrator.

Description

technical field [0001] The invention relates to network measurement, in particular to the application of network anomaly detection and network fault analysis in network measurement, and specifically a network measurement method based on reinforcement learning Sketch. Background technique [0002] Network status detection, network fault analysis, and network security defense are important technologies to ensure the robustness and security of modern networks. Network measurement provides basic information for the above technologies and is the basis of the above technologies. However, with the rapid development of the network, measuring massive amounts of traffic data has become a difficult point in network measurement. It is necessary to improve the accuracy of measurement as much as possible under limited resources. Among them, Sketch is a hash-based approximate data structure, which is widely accepted in the field of network measurement because of its theoretical balance bet...

AI Technical Summary

Problems solved by technology

[0003] However, there is a gap between Sketch’s theoretical guarantee and actual deployment: existing Sketch-based network measurement methods require too many resources in real environments to ensure measurement accuracy, resulting in huge resource overhead; existing Sketch-based network measurement methods in Poor adaptability in changing environments and high performance cannot be guaranteed; existing Sketch-based network measurement tool configuration requires users to have relevant domain knowledge, and network administrators cannot make good use of existing Sketch-based network measurement methods Theoretical configuration, resulting in serious performance loss

Images