传感器网络虚假数据过滤的最优覆盖度分析

ANALYSIS ON OPTIMAL COVERAGE DEGREE FOR FALSE REPORT FILTERING IN SENSOR NETWORKS

虚假数据过滤是传感器网络安全领域的一个重要问题。当前算法通常由t个拥有不同密钥分区的节点协同对数据进行加密,但其所采取的随机部署节点方式无法确保每个事件源都被t个密钥分区所覆盖,从而导致稀疏区域的事件无法上报给sink。提出利用覆盖算法对节点进行部署,基于覆盖质量及网络开销推导和证明了最优节点覆盖度Δ。实验结果表明,采用Δ覆盖比随机部署的密钥覆盖有效性提高约70%。

False report filtering is an essential security issue in security field of sensor networks. Existing algorithms usually encrypt the data report by t nodes with distinct key partitions in collaborative manner; however,the adopted random deployment mechanism cannot guarantee that every area can be covered by t key partitions simultaneously. As a result,the events happening in sparse areas cannot be reported to sink. The paper proposes to deploy nodes through covering algorithm. It then deduces and proves the optimal coverage degree Δbased on covering quality and network overhead. Experimental results show that using Δ coverage can obtain a general increase in efficiency of coverage by about 70% compared with the random deploying method.