| With the wireless sensor network(Abbreviation: WSN)plays a more and more important role in people's lives,information security has become the key to its large-scale and practical,the intrusion detection technology is used in wireless sensor networks,so it will provide more comprehensive and deeper protection for wireless sensor network.This paper mainly focuses on the improvement of clustering routing protocol and the improvement of intrusion detection algorithm in wireless sensor network,because of the limitation of node energy,low utilization of node energy,low intrusion detection rate and poor real-time performance.The main works are as follows:1.This paper adopts hierarchical detection model.Due to the sensor nodes in wireless sensor networks have limited energy and limited computing and storage capacity.This makes the design of the intrusion detection system can not be too complicated,so as not to consume more energy in practical applications,leading to network's performance and lifetime decreased,so the design of intrusion detection model should be considered the characteristics of wireless sensor networks,and should ensure that the correct rate of the intrusion behavior found and real-time.In hierarchical detection model,the common sensor nodes of the perceptual layer are responsible for the collection of data information.The Sink nodes of the aggregation layer is responsible for the work of intrusion feature extraction.The management nodes of the core control layer are used to achieve more complex analysis and detection functions in order to reduce the overall cost of intrusion detection system.2.In this paper,the existing cluster routing protocols are improved.Aiming at the problems of existing clustering routing protocols,such as short network node survival time and low energy utilization,this paper improves the classical LEACH protocol and Fuzzy C-Means protocol based on FCM,and researches a EEUCP protocol.The design idea of this protocol is to optimize the FCM algorithm by using a grey wolf optimization algorithm(Abbreviation: GWO).In addition,the inter cluster communication has also been improved.After comparing to the LEACH protocol,C-FCM protocol and K-Means protocol in the simulation experiment,the results show that the proposed protocol can reduce the energy consumption of the network and prolonging the lifetime of the nodes in the network and can also improve the utilization rate of energy,and make the clustering more uniform.It lays a good foundation for the post intrusion detection work.3.In this paper,the algorithm for intrusion detection in wireless sensor networks is studied in depth.Intrusion detection algorithm is the core of the intrusion detection model.In this paper,the extreme learning machine algorithm(Abbreviation: ELM)with excellent generalization performance is used for intrusion detection.Besides,using principal component analysis(Abbreviation: PCA)for optimization of dimensionality reduction.Then comparing with the traditional back propagation neural network algorithm(Abbreviation: BP).The data set used in the simulation experiment is KDD CUP 99.The results show that the proposed method can greatly improve the detection rate and accuracy,and cut down the false positive rate and false negative rate,and also shorten the detection time.As we can see,the proposed method ensures the accuracy of intrusion detection and real-time. |
