About me

I am a Research Scientist of Meta Platforms.

My current research interests reside in the overlapping area of:

  • Social Media and Quality Systems
  • Mobile Sensing
  • Automobile Systems
  • System Security and User Privacy
  • Internet of Things
  • Wireless Communications and Networking

My full publication list can be found at my Google Scholar Profile..

News

  • Jun. 27th, 2022 I joined Meta as a Research Scientist.
  • Aug. 19th, 2022 I was awarded Ph.D. (Computer Science and Engineering).
  • Oct. 24th, 2019 I presented in MobiCom'19.
  • Dec. 20st, 2017 I became a PhD candicate.


Research Projects

Context-Aware Detection of Vehicle Anomalies

ring

Developed CADD to identify vehicle acceleration anomaly based on context estimations.

  • A novel method, called CADD, for verifying the relationship between control inputs and vehicle dynamics based on estimated context.
  • Mechanisms for inconsistency detection and anomaloussource/group identification.
  • Demonstration of CADD’s performance in detecting data inconsistencies and anomalous source identification via extensive evaluation. CADD is shown to be able to achieve > 97% recall and < 5% false-positive rate with detectiondelay of < 1 cycle. CADD efficiently identifies the anoma-lous group of data with less than 80% accuracy.

CarDog: What’s Wrong with My Car?

ring

Developed CarDog to identify sensor data anomaly of vehicle and external device.

  • Proposed CarDog, a model-based diagnostic system, that can detect an anomaly in vehicle data and also identify the source(s)of anomaly.
  • Evaluated the performance ofCarDogover the real world datacollected from 2 test vehicles and 1 publicly available data set, and demonstrate CarDog’s i) detection performance of individual DSs, ii) ability to narrow down the search space of the anomalysource(s), and iii) end-to-end performance as a diagnostic system.
  • The data-driven approach of CarDog does not require significant modification in vehicle architecture nor the model-specific pa-rameters of the vehicle, thus allowing an easy deployment ofautomated diagnostic system.

Keep Others from Peeking at Your Mobile Device Screen!

ring

Developed HideScreen to protect general on-screen information (e.g., texts and images) shown on our mobile devices.

  • Proposal of grid-based display for on-screen information protection based on optical system properties and human vision characteristics.
  • Development of text protection, HideText, and demonstration of its effectiveness in protecting texts at a low rate of information leakage (≤ 3.8%)
  • Development of image protection, HideImage and SelImage, and demonstration of their effectiveness in hiding images at a low rate of information leakage (≤ 0.9%)
  • Demonstration of HideScreen’s practical usability by evaluating its latency, readability, energy consumption, and users’ feedback
  • Full Paper:
    Chun-Yu (Daniel) Chen, Bo-Yao Lin, Junding Wang, and Kang G. Shin. 2019. Keep Others from Peeking at Your Mobile Device Screen!. In The 25th Annual International Conference on Mobile Computing and Networking (MobiCom '19). ACM, New York, NY, USA, Article 22, 16 pages. DOI: https://doi.org/10.1145/3300061.3300119
    • MobiCom'19 Best Poster Award
    • MobiCom'19 Student Research Competition Second Place
Paper Presentation Slides

In-Car Phone Localization for Detection of Distracted Driving

ring

Developed an in-car phone localization scheme, DAPL (Detection and Alarming of Phone Location), to locate smartphones in moving cars for prevention of distracted driving.

  • Development of a real-time phone localization scheme without requiring any additional ustomized device and direct communication with subsystems in a car.
  • Proposal of an energy-saving mechanism for practical usage scenarios
  • Demonstration of DAPL’s accuracy via extensive experiments in real driving scenarios with 91.77% localization accuracy, 4.23% false positive, and 4.00% false negative rate

Large Scale Reliable Mesh Network

ring

Developed and commercially deployed large scale IoT systems based on IEEE 802.15.4.

  • Implemented a IoT network system from end devices, gateways, to cloud
  • Designed and implemented indoor positioning system, and deployed to multiple commercial sites in Taiwan
  • Designed and implemented secure network entry mechanisms for large scale mesh networks
  • Designed and implemented reliable all-to-one report mechanisms for large scale mesh networks

Green Context-Aware Platform for Smart Living

ring

Developed a context-aware platform that integrates heterogeneous sensors and actuators in smart living environments.

  • Designed a three-layer platform that connects heterogeneous devices with different network technologies
  • Designed automatic configuration mechanisms that enable the devices to set up control links and conduct self-configuration automatically
  • Designed self-powered wireless sensors and radio sniffers to detect user behaviors without interfering the original user habits and intruding user privacies
  • Designed and implemented indoor positioning mechanism based on multi-dimensional scaling (MDS)
  • Designed user-friendly control mechanisms that enable users to control and configure the devices intuitively
Paper

Cooperative localization for wireless and mobile social networking service (SNS)

ring

Developed a cooperative localization scheme that enhances localization accuracy

  • Designed a localization scheme that adopts additional distance information to enhance the localization accuracy of multiple devices
Paper

Publications

Keep Others from Peeking at Your Mobile Device Screen!

ring

People use their mobile devices anywhere and anytime to run various apps, and the information shown on their device screens can be seen by nearby (unauthorized) parties, called shoulder surfers. To mitigate this privacy threat, we have developed HideScreen by utilizing the human vision and optical system properties to hide the users' on-screen information from the shoulder surfers. Specifically, HideScreen discretizes the device screen into grid patterns to neutralize the low-frequency components so that the on-screen information will "blend into'' the background when viewed from the outside of the designed range. We have developed and evaluated several ways of hiding both on-screen texts and images from shoulder surfers. Our extensive experimental evaluation of HideScreen demonstrates its high protection rates (>96% for texts and >99% for images) while providing good user experience.

  • Chun-Yu Chen, Bo-Yao Lin, Junding Wang, and Kang G. Shin
  • Proceeding MobiCom '19 The 25th Annual International Conference on Mobile Computing and Networking
Paper

A Green Context-Aware Platform for Smart Living

ring

Smart living has always been considered as a killer application when new technologies emerged. It is envisioned that smart living will enable a healthier, safer and more comfortable life while reducing resource consumption. Unfortunately, adoption of smart-living technologies and services has been very slow especially in residential environments. Several factors have contributed to such slow and limited deployment. First, installation and configuration of a smart system is usually complicated. Second, maintenance is always a headache for ordinary users. Any repair that requires rewiring is not only time consuming but also very costly. Finally, access to smart-living information and services was not properly addressed. In views of these challenges, a green and context-aware platform is proposed for smart living. In this platform, a full range of self-powered behavioral sensors are developed. These sensors connect objects including toilets, doors, windows, gas stoves, faucets and even tooth brushes to a home network so that various user/environment activities can be recorded. The platform, with the the help of our sensors, is able to provide customized, context-aware services without using any intrusive sensors such as cameras. The platform is implemented in an off-campus apartment to demonstrate its potentials.

  • Chun-Yu Chen, Yu-Jen Ku, Chih-Wei Ho, Yan-Ze Lin, and Chun-Ting Chou
  • IEEE 6th International Conference on Service-Oriented Computing and Applications, December 2013
Paper

Cooperative localization for wireless and mobile social networking service (SNS)

ring

Wireless and mobile social networking service (SNS) utilizes location information to enrich user experience. A major challenge in location-aware SNS is its strict requirement in precision of user location, which generally cannot be met by the existing GPS or cellular networks. In this paper, we propose a user-level cooperative localization scheme that improves the precision of existing localization techniques. A mathematical model is developed and in-depth simulation is conducted to evaluate the performance of the proposed scheme.

  • Yu-Chung Chen, Chun-Ting Chou, and Chun-Yu Chen
  • IEEE International Wireless Communications and Mobile Computing Conference (IWCMC), July 2011
Paper