适用于虚假数据过滤的最优传感器覆盖研究*

已有传感器网络虚假数据过滤机制采用随机策略部署节点，由转发节点对数据包中附带的t个MAC(Message Authentication Code)签名进行验证，从而实现对虚假数据的识别和过滤。然而，在实际应用中，随机部署往往在网络中形成部分稀疏区域，无法被t个拥有不同密钥分区的节点同时覆盖。提出利用覆盖算法对节点进行部署，在均衡覆盖质量及网络开销的情况下，证明了适用于虚假数据过滤的最优节点覆盖度为2t，并进一步推导了一些相关的覆盖结论。理论分析及仿真实验表明，与随机部署相比，最优覆盖算法极大提高t个密钥分区同时覆盖的概率。例如，当400个节点部署于50?50m2的区域时，随机部署和最优覆盖算法保证t个密钥分区同时覆盖的概率分别为9%和92%。

Research on optimal sensor coverage for false data filtering

In wireless sensor networks, existing false data filtering mechanisms exploit random method to deploy nodes, and verify the attached t MACs (Message Authentication Codes) in each data report on intermediate nodes, thus to detect and filter out false reports. However, in actually applications, random deployment method usually forms some sparse areas in the network which cannot be covered by t nodes with distinct key partitions simultaneously. With an aim to balance covering quality and overhead of the network, this paper first analyzed and proved the optimal covering degree 2t for false data filtering, and then exploited the corresponding optimal covering algorithm to deploy nodes. Moreover, it also derived some useful covering theorems. Theoretical analysis and simulation results demonstrate that, compared with random deployment, optimal covering can improve the efficiency of covering by t keys simultaneously. For example, when there are 400 nodes deployed in an area of 50?50m2, the probabilities of random deployment and optimal covering algorithm to guarantee covering by t keys simultaneously are 9% and 92%, respectively.