An Intrusion Detection Method for Wireless Sensor Networks Based on Ensemble Learning

Abstract

The invention proposes an integrated learning-based wireless sensor network intrusion detection method, which belongs to the technical field of communication. The method includes: collecting the data of each node from the wireless sensor network, preprocessing the data, extracting the feature set of each node, converting symbolic features into numerical values; normalizing each feature value; using feature selection The algorithm selects the optimal feature set from the preprocessed feature set; uses the improved SVM algorithm as a weak classifier, and combines the selected optimal feature training set for training; uses the Adaboost ensemble learning algorithm to combine the trained weak classifiers Then use the trained strong classifier to test the actual data to distinguish the normal nodes and abnormal nodes of the wireless sensor network. The method can improve the accuracy of intrusion attack detection in the wireless sensor network, reduce the cost of certain marked samples, reduce intrusion detection detection and training time, and enhance the reliability of the intrusion detection system.

Description

technical field [0001] The invention belongs to the technical field of communication, and relates to an intrusion detection method of a wireless sensor network based on integrated learning. Background technique [0002] Wireless Sensor Networks (WSN) is a distributed sensing detection system composed of many sensor nodes deployed in a wide area. The sensor nodes perceive and monitor the information in the deployment area, and transmit data through multi-hop routing. to the cluster head, and forwarded to the base station through the cluster head or forwarded to the base station through other cluster heads to realize data collection and task detection. Due to the characteristics of low node cost, simple deployment, no need for infrastructure, strong invulnerability, and strong adaptability to highly dynamic network topologies, wireless sensor networks are widely used in vehicle monitoring, environmental monitoring, intelligent transportation, intelligent medical care, intellig...

AI Technical Summary

Problems solved by technology

[0004] Since the sensor node deployment algorithm will be limited by the computing power of the node, if you use algorithms and models that are computationally complex and require a large amount of data for training, it will block the normal work of SN or CH, increase the performance of SN or CH, and cause SN or CH inventory cycle reduction

Images