Pedestrian distraction continues to be a significant pedestrian safety concern in urban communities. Existing pedestrian safety frameworks and applications are unable to simultaneously detect complex distraction-related activities (often focusing on detecting only specific activities and contexts) and ignore the hazards posed by the distracted pedestrian to fellow pedestrians and drivers. Moreover, a majority of existing complex activity recognition schemes are either computationally impractical for real-time implementation on mainstream mobile and wearable devices, or employ specialized auxiliary hardware, or both. Thus, there is an urgent need for usable and accurate pedestrian safety solutions that can be efficiently implemented (and used) on commercial off-the- shelf (or COTS) mobile and wearable devices. This project proposes a cloud-assisted pedestrian safety framework that detects several commonly observed activities resulting in pedestrian distraction by using multi-modal and multi-source data from users’ mobile and wearable devices, and provides appropriate on-device and community notifications, with the objective of achieving a favorable balance between responsiveness, computational efficiency, detection accuracy and usability.
Smart wearable devices, such as smart watches, are very popular and fast replacing their traditional non-smart counterparts. By means of various high-precision on-board sensors, these devices capture rich contextual information about the wearer and his environment to enable several new and useful applications. However, this diverse set of on-board sensors also provides an additional attack surface. Access to these sensors, if not controlled appropriately, can be used as a side-channel by an adversary keen on obtaining private and sensitive information belonging to the wearer. Moreover, active misuse detection and resistance of these wearable device sensors is not straightforward. There is currently a lack of understanding of the various side-channel security vulnerabilities that are possible due to wearable devices and there is an urgent need to study the means for continuously protecting against them. The research in this project addresses this very timely topic.
The goal of this research is twofold: first, to demonstrate that wearable devices enable novel side-channel security and privacy threats, and second, to design continuous authentication techniques and adaptive access control mechanisms to survive these threats. Specifically, this research will evaluate private data inference and wearer tracking threats in wearable devices that utilize unprotected sensors as side-channels. This will be accomplished by designing appropriate learning-based classification and prediction mechanisms that can be used by an adversary for inferring sensitive data. On the protection front, this project will develop a multi-sensor activity and identity classification framework. This framework will leverage rich contextual sensor data (e.g., fine-grained movements, application usage and critical body parameters) to enable continuous identification and authentication of legitimate wearers and their activities.
Significant developments in the electric power industry are in the areas of advanced measurements, improved communication infrastructure, renewable energy sources, and electric vehicles. These changes are expected to influence the way energy is provided to and consumed by customers. Advanced Metering Infrastructure (AMI) initiatives are a popular tool to incorporate these changes for modernizing the electricity grid, reduce peak loads, and meet energy-efficiency targets; however, privacy concerns have limited customer acceptance of these initiatives. The research objective of this project is to design appropriate architectures for information collection and dissemination with security and privacy guarantees and to develop state-of-the-art algorithms and protocols for privacy-preserving communication and control that effectively exploit the AMI for improved system operations and active customer participation.
The increasing popularity of Online Social Networks (OSNs) is spawning new security and privacy concerns. Currently, a majority of OSNs offer very naive access controls that are primarily based on static Access Control Lists (ACL) or policies. But as the number of social connections grow, a static ACL based approach slowly becomes ineffective and unappealing to OSN users. There is an increased need to control access to data based on the associated context, rather than solely on data ownership and social links. Surveillance by the OSN service provider is another critical concern for OSN users, as the service provider may further scrutinize data posted or shared by users for personal gains (e.g., targeted advertisements), for use by corporate partners or to comply with legal orders. In this project, we introduce a novel paradigm of context-based access control in OSNs, where users (in the sharer’s social network) are able to access the shared data only if they have knowledge of the context associated with it.
Secure localization in the presence of cheating or non-trustworthy anchors is an important problem in wireless networks. Although many secure localization schemes have been proposed in the past, efficiently eliminating the cheating effect of malicious anchors in such localization schemes still remains an open problem. In this project, our goal is to address this open problem. One of the solutions that we propose employs a novel Code Division Multiple Access (CDMA) based communication protocol in order to overcome the problem of cheating anchors.