Aware-data-based wireless sensor network abnormity type discriminating method

Abstract

The invention provides an aware-data-based wireless sensor network abnormity type discriminating method, which relates to the security field of wireless sensor network information. According to the method of the invention, the wireless sensor network nodes collect a group of aware data every a fixed interval and transmit the data to a base station. In a normal time period, the detection characteristic set generated by the base station is used as a training set and the training set is normalized with the column mean value and the column variance unchanged. The training set undergoes dimension reduction through a main component analyzing method and the characteristic vector matrix and the column mean value vector are kept. The clustering of the training set is divided into normal clustering and abnormal clustering through a density-based competitive clustering algorithm. When new detection characteristics arise, based on the normalization of the column mean value and the column variance and through the dimension reduction of the characteristic vector matrix and the column mean value vector, it is possible to detect whether the network is abnormal or not according to the gains of the normal clustering and the abnormal clustering the network is divided into. The deployment of the method is simple and is low in cost. Despite that, the method can detect network protocol attacks and malicious data injection attacks, and can reduce the energy of the nodes.

Description

technical field [0001] The invention belongs to the field of wireless sensor network information security, and relates to a method for identifying abnormal types of wireless sensor networks based on perception data. Background technique [0002] Wireless sensor network is a distributed sensing system composed of a large number of micro sensor nodes. The system can collect environmental information (sensing data) in the monitoring area in real time, such as temperature, humidity, brightness and pressure, etc., and transmit the sensing data to the base station in a multi-hop wireless manner. Wireless sensor networks are usually deployed in unattended, harsh environments, or even in hostile areas, where attackers can easily capture sensor nodes and invade the network. In addition, sensor nodes are limited in energy, communication capabilities, and computing and storage, making them extremely vulnerable to various attacks. Intrusion Detection System (IDS) is an important secur...

AI Technical Summary

Problems solved by technology

The simulation results of this algorithm show that the successful defense rate of this algorithm is much higher than the scenario where game theory and Markov decision process are used alone. The disadvantage is that frequent communication between nodes in the game process will lead to excessive energy consumption.

Although the author proposes that strategies such as low-power routing protocol (LEACH), random selection of cluster heads, and high-efficiency energy hierarchical clustering can be used to reduce the communication overhead of nodes, nodes still need to transmit messages such as identification and verification for intrusion detection activities, and The paper does not give the experimental results of the relationship between the intrusion detection algorithm and the energy consumption of nodes

The energy consumption of sensor nodes is mainly in the three aspects of sensing the surrounding environment, processing sensory information and transmitting data (data includes sensory data and network protocol data), among which the energy consumption of nodes is mainly in transmitting data, and intrusion detection systems usually require nodes to record And transmit characteristic information for detection, which will increase the energy consumption of nodes for intrusion detection activities

Images