Wang Na, Wang Jiacun, Chen Xuemin
Department of Computer and Information Engineering, Shanghai Polytechnic University, Shanghai 201209, China.
Department of Computer Science and Software Engineering, Monmouth University, West Long Branch, NJ 07764, USA.
Sensors (Basel). 2019 Apr 23;19(8):1916. doi: 10.3390/s19081916.
Wireless Sensor Networks (WSNs) are prone to failures and malicious attacks. Trust evaluation is becoming a new method for fault detection in WSNs. In our previous work, a comprehensive trust model based on multi-factors was introduced for fault detection. This model was validated by simulating. However, it needs to be redeployed when adjustment to network parameters is made. To address the redeployment issue, we propose a Trust-based Formal Model (TFM) that can describe the fault detection process and check faults without simulating and running a WSN. This model derives from Petri nets with the characteristics of time, weight, and threshold. Basic structures of TFM are presented with which compound structures for general purposes can be built. The transition firing and marking updating rules are both defined for further system analysis. An efficient TFM analysis algorithm is developed for structured detection models. When trust factor values, firing time, weights, and thresholds are loaded, precise assessment of the node can be obtained. Finally, we implement TFM with the Generic Modeling Environment (GME). With an example, we illustrate that TFM can efficiently describe the fault detection process and specify faults in advance for WSNs.
无线传感器网络（WSN）容易出现故障和遭受恶意攻击。信任评估正成为无线传感器网络中故障检测的一种新方法。在我们之前的工作中，引入了一种基于多因素的综合信任模型用于故障检测。该模型通过模拟进行了验证。然而，当对网络参数进行调整时，它需要重新部署。为了解决重新部署问题，我们提出了一种基于信任的形式化模型（TFM），该模型可以描述故障检测过程并在不模拟和运行无线传感器网络的情况下检查故障。这个模型源自具有时间、权重和阈值特征的Petri网。给出了TFM的基本结构，利用这些基本结构可以构建通用的复合结构。为了进一步进行系统分析，还定义了转移触发和标记更新规则。针对结构化检测模型开发了一种高效的TFM分析算法。当加载信任因子值、触发时间、权重和阈值时，可以获得对节点的精确评估。最后，我们使用通用建模环境（GME）实现了TFM。通过一个例子，我们说明了TFM可以有效地描述故障检测过程，并为无线传感器网络预先指定故障。
