Reputation-aware Trajectory-based Data Mining in the Internet of Things (IoT)

Speaker Bio

Samia Tasnim is a Ph.D. candidate at the School of Computing and Information Sciences (SCIS), Florida International University (FIU). She is working under the supervision of Dr. S. S. Iyengar and Dr. Niki Pissinou. Samia received her Bachelor’s degree in Computer Science and Engineering from Bangladesh University of Engineering and Technology (BUET) and Masters in Computer Science from FIU in 2015. Her research interests include internet of things, mobile computing, and data mining. Samia worked as a software engineer for two years before starting graduate studies at FIU. After her graduation, she will join Florida A&M University as a tenure-track assistant professor.

Description

Internet of Things (IoT) is a critically important technology for the acquisition of spatiotemporally dense data in diverse applications, ranging from environmental monitoring to surveillance systems. Such data helps us improve our transportation systems, monitor our air quality and the spread of diseases, respond to natural disasters, and a bevy of other applications. However, IoT sensor data is error-prone due to a number of reasons: sensors may be deployed in hazardous environments, may deplete their energy resources, have mechanical faults, or maybe become the targets of malicious attacks by adversaries. While previous research has attempted to improve the quality of the IoT data, they are limited in terms of better realization of the sensing context and resiliency against malicious attackers in real time. For instance, the data fusion techniques, which process the data in batches, cannot be applied to time-critical applications as they take a long time to respond. Furthermore, context-awareness allows us to examine the sensing environment and react to environmental changes. While previous research has considered geographical context, no related contemporary work has studied how a variety of sensor context (e.g., terrain elevation, wind speed, and user movement during sensing) can be used along with spatiotemporal relationships for online data prediction.

This dissertation aims at developing online methods for data prediction by fusing spatiotemporal and contextual relationships among the participating resource-constrained mobile IoT devices (e.g. smartphones, smart watches, and fitness tracking devices). To achieve this goal, we first introduce a data prediction mechanism that considers the spatiotemporal and contextual relationship among the sensors. Second, we develop a real-time outlier detection approach stemming from a window-based sub-trajectory clustering method for finding behavioral movement similarity in terms of space, time, direction, and location semantics. We relax the prior assumption of cooperative sensors in the concluding section. Finally, we develop a reputation-aware context-based data fusion mechanism by exploiting inter sensor-category correlations. On one hand, this method is capable of defending against false data injection by differentiating malicious and honest participants based on their reported data in real time. On the other hand, this mechanism yields a lower data prediction error rate.