Secured Data Collection for a Cloud-Enabled Structural Health Monitoring System

Sensing data collected from some applications such as structural health monitoring is complex. When collected at a high-rate and -resolution, it is big data. A wireless sensor can neither store/process all raw data locally nor reliably forward the data. When attempting to achieve application monitoring from a remote monitoring center using Cloud, challenge appears in terms of quality of monitoring. However, the sensing data provided by a sensor or a group of sensors (or a cluster) are usually not secured; the data are often compromised by security attacks. Therefore, it becomes difficult to identify the compromised data once the data reaches Cloud storage. In this paper, we propose a novel cloud-enabled remote structural health monitoring (cSHM) framework, by which aggregation and monitoring decision can be made based on secured data stored on Cloud. We mainly focus on preparing the secured data for uploading to the cloud, which is carried out at network level before uploading. To achieve this, we present a truth discovery approach, whose goal is to infer truthful facts from insecure sensor signals. If any single signal or a set of signals is compromised, the framework attempts to recover it by a signal reconstruction algorithm. We perform experiments through real data traces and demonstrate that data forwarding to the cloud from the network level is secured and protected so that a highquality decision on an structural event event detection can made based on the data stored on Cloud.