A GM model-based traffic anomaly detection method for wireless sensor networks

Abstract

The invention discloses a GM model-based wireless sensor network traffic anomaly detection method; using the GM (1,1) model, it has the characteristics of less historical data, fast model building speed, and accurate prediction value, and is very suitable for wireless sensor network nodes Conditions with limited energy and computing power; use a sliding window with an appropriate window size to fix the amount of historical modeling data, which not only ensures the speed of modeling, but also ensures the latest validity of historical data; optimizes GM (1,1 ) model's whitening differential equation to solve the initial conditions, so that the predicted value is more accurate; the flow predicted value at the next moment, which is finally used for abnormal judgment, is generated by the exponential weighted average of the previous L predicted values, so that the flow prediction introduces a certain " Inertia", when abnormal traffic comes, the normal traffic prediction model cannot be easily changed, but the predicted value of normal traffic can be better obtained, and abnormal traffic can be detected more easily.

Description

technical field [0001] The invention belongs to the technical field of network security, and in particular relates to a GM model-based abnormality detection method for wireless sensor network flow. Background technique [0002] With the development of communication and computer technology, the network has become an important factor in the development of today's world. As one of the important network technologies, wireless sensor networks (Wireless Sensor Networks) are widely used in national defense, military and national security due to their advantages of high robustness, high accuracy, high flexibility and strong intelligence. , environmental monitoring, traffic management, medical and health, manufacturing, anti-terrorism and disaster relief, and other fields are also the main way for the Internet of Things to obtain information. The wireless sensor network can sense and collect information of various environments or monitoring objects through the real-time monitoring o...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the defect of relatively large algorithm complexity that is common in current wireless sensor network traffic anomaly detection methods. In order to achieve more efficient real-time traffic anomaly detection under the premise of ensuring detection accuracy, a new method is proposed. A GM model-based traffic anomaly detection method for wireless sensor networks

Images