Research Projects


Active Projects


ProperData logo

ProperData: Protecting Personal Data Flow on the Internet

Project Page:  https://properdata.eng.uci.edu/

Personal data collection typically starts on user devices, in a range of application domains (web, mobile, IoT). Data are then shared with service providers as well as with a large number of trackers. Data can also be obtained by malicious actors and/or used for surveillance. Many useful services are enabled by the collection of this data, although increasingly at the expense of privacy, security, transparency, and fairness, for individuals and society as a whole. Increased public awareness has led to recent legislation on data protection, such as GDPR and CCPA, and policy has become a powerful tool to be used in synergy with technology. This NSF SATC Frontier project seeks to protect personal data, by improving the transparency and control of personal data flow on the Internet. We take a multidisciplinary approach, combining methodologies from computer science and engineering (theory, network measurement, systems, security) with policy and concepts from economics. For more information, please see the ProperData website and please see the project-specific links below:


Network Traffic Filtering against Advertising and Tracking

Project Page: Network Traffic Filtering 

Summary: In this project, we collect and analyze network traffic generated by different types of devices and platforms at the edge of the network, including: mobile devices, smart TVs, web browsers and IoT devices. Our goal is to understand and eventually defend against advertising and tracking services (a.k.a. “ATS”). We extract features from multiple layers (at least from the network and often from applications as well) and we train machine learning models that can used for real-time detection and blocking of advertising and tracking, anomaly detection, and other purposes. In particular, we collect and analyze network traffic generated by the following types of end-devices and platforms:


Private-Preserving Mobile Data Crowdsourcing

Project Page: Privacy-Preserving Mobile Data Crowdsourcing

Summary: A large number of services (such as wireless connectivity, location-based services, recommendations, etc.) rely on data and measurements collected from mobile devices and shared with a centralized entity. Users enjoy these services at the expense of sharing data from their mobile devices, which increases their privacy risk. For example, users may want to receive a location-based service, but they may be concerned about sending their exact location to a server or third parties. In this project, we consider data collected from mobile devices, including information about the wireless network itself, as well as personal user data. Our goal is to develop privacy-preserving techniques to obfuscate reported data, while still providing guarantees for the quality of the provided service.


N-Body Algorithms for Mobile and Social Data

Project Website: NSF EAGER 

Summary: Many problems in the above space can be framed in terms of pairwise interactions among spatially embedded entities,  traditionally the domain of N-body problems  in the physical sciences. Therefore, revisiting and adapting N-body algorithms specifically to mobile and social data  analysis and learning  can increase our capacity to (i) work with such data at scale and (ii) do so in a privacy-preserving way.  We investigate novel N-body and parallel algorithms, specifically designed for processing hierarchical, and geospatially embedded, mobile and social data, where the size or access to datasets is prohibitive. In addition to the algorithmic design, we develop software modules (e.g.,  special purpose compilers, crowdsourcing tools and generators of synthetic datasets) that implement these methods. Application domains include: cellular network monitoring, mobile data privacy, social network modeling, statistical network analysis, web analysis, scalability and performance analysis of mobile devices.


Past Projects


Network Sampling and Construction

Project Page: III

Summary:  The goal of this project is to study network data that are generated in the context of mobile and/or online social networks. The project develops methods for (i) network sampling to facilitate inference for network structure and/or attributes and, conversely, for (ii) construction of networks with target characteristics. The methods aim at improving the state-of-the-art in network inference and network data anonymization, with target application domains primarily mobile and social network data. Here is the NSF Abstract.


AntMonitor

Project Page: AntMonitor

Summary: AntMonitor is mobile software that runs on the mobile device, and passively monitors all packets in and out of the network interface. We designed AntMonitor as a VPN-based service, and we developed and compared two versions of the architecture: Client-Server and Mobile-Only. We demonstrated  the lean performance of the AntMonitor Mobile-Only prototype, in terms of throughput and energy, and compared it to  the Client-Server one, as well as other state-of-the-art VPN-based approaches. For example, it achieves speeds of over 90 Mbps (downlink) and 65 Mbps (uplink), which are 2x and 8x throughput of existing mobile-only approaches, and at 94% of the throughput without VPN,  while using 2–12x less energy. AntMonitor can be used as a tool to support a number of passive monitoring applications, including: real-time detection and prevention of private information leakage from the device to the network; packet classification to predict a number of properties including ads, applications, etc based on packet headers; and passive performance measurements.

Online Social Networks

Project Page: Online Social Networks