Wei Dong            董玮

PhD, Professor, PhD supervisor
College of Computer Science, Zhejiang University

Email: dongw AT zju.edu.cn
Office:
 
R.m. 311, Zetong Building, Yuquan Campus, Zhejiang University, Hangzhou, China.
News Projects Publications Grants Awards Talks Misc

Biography

Wei Dong is currently a full Professor at the College of Computer Science and Technology in Zhejiang University. He received the BS degree and PhD degree from the College of Computer Science and Technology in Zhejiang University in 2005 and 2010, respectively. He was a Postdoc Fellow at the Department of Computer Science and Engineering in Hong Kong University of Science and Technology in 2011. He joined in Zhejiang University as a faculty member in Feb., 2012. He leads the Emerging Networked Systems research group (EmNets). His current research interests include AIoT, edge computing and edge AI, wireless networking and IoT security.

News

Recent Projects

Confidential sensing computation for IoT: Confidential sensing computation, which applies confidential computing to the process of sensing, computation, and inference of IoT data, is an important technique to ensure the secure use of data and code on IoT devices. We conduct various researches on this topic, including programming methods for TEE-enabled IoT devices, lightweight isolation methods like WebAssembly, and AI inference with data privacy protection and model privacy protection, etc. We have proposed some novel methods and techniques include dTEE [IPSN'24], FDAS [IMWUT/UbiComp'24], WAIT [MobiSys'22], and etc.
AI-empowered IoT: The rapid growth of AI technologies, e.g., large language models, will have a great impact of IoT. We are especially interested in how to embed intelligence into a vast number of IoT devices. We study how to tailor complex AI models for resource-contrained IoT and edge devices, how to use AI for intelligent sensing, how to use AI to facilitate IoT programming, and how to use AI to empower various IoT applications such as video surveilance, smart home, and mobile computing. We have proposed some novel techniques and systems include AirText [TMC'23], EINet [ICDCS'23], MEEdge [WWW'24], ChatIoT [IMWUT/UbiComp'24]
Low-power wireless: Recent years have witnessed many low-power wireless technologies, e.g., ZigBee, Bluetooth, LoRa, etc. These heterogeneous technologies raised several important research issues, such as how to combat against cross-technology interference, how to perform cross-technology communication without a gateway, how to interoperate over a wide range of IoT protocols. In this study, we aim to address several key issues in this field. We have proposed some novel techniques and systems include TinyNet [MobiSys'22], LoFi [INFOCOM'21], RT-BLE [INFOCOM'23], BTrack [INFOCOM'24], BLEW [SenSys'24] and etc.
Low-code IoT programming: Low-code IoT programming is very important for simplify complex IoT development process. We conduct researches on various issues on this topic, including how to adapt to various IoT hardwares, how to facilitate device-edge-cloud integrated programming, and how to embrace LLM for IoT development. Some novel systems include TinyLink [MobiCom'17, MobiCom'20], LinkLab [IoTDI'19, NSDI'23], EdgeProg [ICDCS'21, TC'22], WiProg [INFOCOM'21].
Novel IoT systems and applications: We have built various realitics IoT systems and applications. These include the GreenOrbs system for forest monitoring, the Mosaic mobile system air quality monitoring, and the IoTbridge system for vibration monitoring for the Hong Kong-Zhuhai-Macao bridge. We conduct research on system measurement, design and deployment issues. Some research papers include measurement on GreenOrbs [INFOCOM'13, ToN'14], Mosaic [INFOCOM'16, INFOCOM'17, UbiComp'18], etc.

Selected Publications (full list)

Grants

Current grants

Past grants

Awards

Invited Talks

Misc