Quick Links
Sponsors - NSF CSR Grant.
Research - Other On-Going Research.
Publications - Related Publications
People - Team Members
Rapid advancement of the wireless technologies provide new opportunities for mobile users to have easy access to real-time data, derive useful social information, and stay connected with business partners, colleagues and friends. Towards this end, mobile social networking applications have recently emerged to meet these needs. Current mobile social networking applications do not support advanced context-based services. Additionally, serious security and privacy concerns have been raised when accessing social networking applications either from fixed locations or on-the-go. This project aims to build a secure mobile information sharing system (SEMOIS) that supports secure and privacy-preserving real-time information sharing. SEMOIS has the ability to store secure data items with flexible access control at insecure storage nodes and enables users to send context-based messages with late-binding features. SEMOIS achieves data confidentiality and privacy-preserving through data encryption and encrypted search, and enables intentional name based message dissemination without apriori knowledge of recipients. SEMOIS will only focus on intradomain operations i.e. a particular data schema is used for all considered nodes. Another related project will study the interdomain issue. Additionally, a set of smart learning methods are developed to extract short-term and long-term geo-social patterns from multimodal sensing data collected by mobile devices for social networking purposes, e.g., geo-social patterns are used to derive hidden communities.
Project results are expected to advance the state of the art techniques for supporting secure and privacy-preserving mobile social networks with a variety of innovative features. The project equips both graduate and undergraduate students with the necessary background and practical skills for survival in the emerging job market and further contributes to the development of the pervasive computing field. In addition, SEMOIS can be used by middle and high school students from the Tri-State area that participate in the CHOICES and NSF-funded STEM program organized by Lehigh University.
We have designed a scheme that utilizes social community information to control the propagation of diseases in HealthCare domain. Interested readers can find more details from our technical report
Mobile devices play important roles in our daily life. Users store information in their mobile devices and would like to share such information with others via opportunistic peer-to-peer links. However, such opportunistic links are intermittent in nature and hence require the store-and-forward feature proposed in delay tolerant networks to provide useful data sharing opportunities. Due to the limited resources, users with mobile devices may not be willing to help forward data items belonging to others. Thus, we design a multi-receiver incentive-based dissemination scheme (MuRiS) which allows nodes to cooperatively deliver information of interest to one another via chosen delivery paths which utilize few transmissions. Our results were published in an IEEE Globecom 2012 paper and a journal paper which appeared at IEEE Transaction in Wireless Communcations, Vol 13, Issue 1, 2014.
We have designed a low cost monitoring system using wearable sensors equipped with accelerometers and audio. Behavior dtection scheme relies on robust data mining techniques. Historical traces are mined to build individuals' activity and environmental classifiers as well as construct a stressful environment table which is then used to generate alerts to caregivers. Evalution of this system with one teenage boy with stereotypical behavior does not reveal any association with environmental trigger. However, we do notice the change in its frequency whenever his medication changes.
We have also developed two useful mobile applications: (a) conversational trainer application, (b) story game. Our conversational trainer application was used by a speech therapist at Centennial School to improve conversational skills of special need children. The applications have been released to the public via the following link: http://carina.cse.lehigh.edu/graceland.
Mobile users often like to use images taken by their smartphone to gather more information. We designed an efficient mobile visual search system called EMOVIS for landmark recognition. Our EMOVIS paper is published in IEEE MSN 2013. In addition, we have also designed an efficient on-device mobile visual search system. Our design is based on the Bag-of-Visual Word framework but uses small visual dictionary. An Object Word Ranking algorithm is proposed to efficiently identify the most useful visual words of an image to construct compact image signature for fast retrieval. We also design a more efficient ranking-consistency re-ranking algorithm to improve the retrieval accuracy. We demonstrate the effectivness of our design via our prototype EMOD system for a database with over 10K images. Our results are documented in this conference paper at ACM Multimedia System Conference in 2015.
Furthermore, we design a low cost gait-based user authentication scheme for healthcare monitoring system. Our solution involves identifying walking cycles and performing normalization such that our verification scheme is robust against users' varying walking speeds. Our framework can be deployed in user-centric or server-centric manners. Extensive evaluation using 3000 smartphone-based traces show that our scheme is very robust. Our results have been documented in an IEEE SECON 2013 paper, and an accepted journal paper which will appear in IEEE Transactions on Mobile Computing.
Sponsors - NSF CSR Grant.
Research - Other On-Going Research.
Publications - Related Publications
People - Team Members