一种检测被捕获节点的基于异常的算法研究

发布时间：2018-05-28 07:35

本文选题：无线传感器网络 + 捕获　；参考：《压电与声光》2014年06期

【摘要】：无线传感器网络在许多应用场合里需要采集较敏感的数据,因此安全问题至关重要。一旦传感器节点被捕获,且没有采取相应措施,节点的密钥信息易被泄露,攻击者完全可伪装成这些节点,向网络任意注入错误的信息,由此导致网络的安全性能急剧下降。提出了针对被捕获节点的一种基于异常的入侵检测算法,能有效识别无线传感器网络的被捕获节点。算法对传感器节点间关系进行抽象,采用传感器网络的事件驱动特性来确定某节点在固定时间间隔内是否在发生数据包,基站通过检测可疑节点的数据包发送时间的差异来加以确认。算法不依赖于任何被捕获节点如何行动和密谋的假设,能识别出偏离正常行为值的最大多数被捕获节点,而不会出现"假肯定"。
[Abstract]:Wireless sensor networks need to collect sensitive data in many applications, so the security problem is very important. Once the sensor nodes are captured and no corresponding measures are taken, the key information of the nodes is easy to be leaked. The attacker can completely disguise these nodes and inject any wrong information into the network, which leads to a sharp decline in the security performance of the network. An anomaly based intrusion detection algorithm for captured nodes is proposed, which can effectively identify the captured nodes in wireless sensor networks. The algorithm abstracts the relationship between sensor nodes and uses the event-driven characteristics of sensor networks to determine whether a node is generating packets at a fixed time interval. The base station verifies by detecting the difference in the packet transmission time of the suspect node. The algorithm does not depend on any assumption of how the captured node acts and conspires, and it can identify the most captured nodes that deviate from the normal behavior value, without "false affirmation".
【作者单位】：湖北第二师范学院物理与机电工程学院;Florida
【基金】：2012年湖北省教育厅科研计划基金资助项目(B20123102)
【分类号】：TN929.5;TP212.9

【参考文献】