DDoS Attack Detection Model Based On Trust And Authorization In Internet Of Things

With the continuous development of the Internet technology and sensor technology,the Internet of things had a rapid development.At the same time,it also exposed many urgent security problems need to be solved.Mostly Internet of things had low computing power devices and limited resources,most of which were in unsupervised areas.They were vulnerable to attack or controlled as members of bot-net which led to launch large-scale DDoS attacks.In addition,the Internet of things had huge scale and the characteristics of multiple network convergence made the traditional network security strategy had difficulty to implement.How to formulate a targeted and effective security strategy to resist the DDoS attacks of Internet of things was a urgent key issue needed to be solved in the continuous development of the Internet of things.Trust assessment model could perfectly adapt to the characteristics of large scale and multiple network fusion of the Internet of things.In order to solve the lack of pertinence of the current trust assessment model to the Internet of things environment that is the simple node had difficulty to use the same evaluation calculation as high computing resources node to present a trust assessment model based on trust authorization and authorize the two sides of the interaction separately.And cluster structure was used to transfer the computing task to the cluster head node.The simulation experiment showed that the success rate of interaction is improved obviously and it had the resist ability in a degree to the malicious recommendation.The simple Internet of things nodes were vulnerable to DDoS attacks or were controlled as zombie nodes.This article proposed a LVQ DDoS testing mechanism based on trust,which tested the interactive traffic anomaly in both proactive and passive situations.On the one hand,the introduction of trust side selectively reduced the sampling period of the higher trust node;on the other hand,the probability of the high trust node being controlled was smaller than the low trust node.The simulation experiment showed that the model had higher rate DDoS of attack detection and lower rate of false detection.