Research
Projects

• Cross-Language Evaluation of IoT Device Drivers for Security Verification Using LangChain
• Create a local vector database using Redis to store embeddings of device driver code written in different programming languages that are generated by OpenAI Embedding model.
• Build a RAG model on a local Hugging Face LLM within LangChain framework to retrieve relevant code in sequential queries with well-designed and structured prompts. Employ an Output Parser to format each response for the subsequent query.
• Build an LLM-based Agent equipped with various tools to discover discrepancies in the implementation of device driver code on different IoT platforms for security analysis.

• Discovering and Exploiting Vulnerability on Zigbee Devices
• Discovered and demystified a vulnerability - Zigbee Hidden Attributes on widely-used Zigbee devices and implemented an end-to-end attack to exploit the vulnerability. Built an attack device using nRF5 SDK in C language on an nRF52840 DK.
• Timely disclosed this vulnerability to device manufacturers, including Philips, Samsung, Amazon, Frient, Yale (lock manufacturer), Kwikset, and Connectivity Standards Alliance (CSA).
• Received acknowledgements from Samsung, Amazon, and CSA. Furthermore, Amazon awarded a high bounty of $2,500 (Rank: 16) via HackerOne for the vulnerability disclosure. Check it out here!
• The paper Discovering and Exploiting IoT Device Hidden Attributes: A New Vulnerability in Smart Homes has been accepted by ACM CCS 2025. [PDF]

• Defensive System against the Delay Attacks in Smart Homes
• Built an one-stop-for-all system in JavaScript in order to 1) connect and manage various IoT devices with different connectivities (e.g., Zigbee, Z-Wave, and Wi-Fi); 2) integrate with two popular IoT cloud platforms (i.e., Samsung SmartThings and IFTTT) with OAuth 2.0 implemented for authentication. A database was implemented using SQLite to store device status, as well as access tokens for interfacing with the cloud platforms.
• Proposed an effective timeout-based approach to detect the delay attacks and then creatively used OpenVPN to handle the detected delay attacks.
• The paper MP-Mediator: Detecting and Handling the Stealthy IoT Event and Command Delay Attacks has been published in RAID 2023. [PDF]

• Detection of Malicious Local Attacks on IoT Devices
• Built an OpenWrt Wi-Fi router on a Raspberry Pi and adopted tcpdump to remotely capture network traffic for further analysis. Used Python scripts to analyze network layer information of the captured packets to detect malicious local attacks that exploit device local APIs, based on communication patterns between various parties in smart home IoT systems.
• Simulated malicious local attacks by developing an iOS app in Swift to send malicious commands to the victim IoT devices, including plugs, light bulbs, lock, and vacuum.
• The paper IoTtracer: Detecting Malicious Local Attacks on IoT Devices by Utilizing IoT System-level Traffic Patterns has been submitted to one of the top-tier four security conferences.

• End-to-End Smart Speaker Protection System
• Developed an Android app with Firebase Cloud Messaging integrated to measure the proximity of the user to the smart speaker without manual operation when the smart speaker is invoked, even if the smartphone is in the pocket all the time.
• Built an end-to-end protection system in Python to first detect unauthorized voice commands using traffic analysis and then block them from being sent to the cloud servers for further processing.
• The paper VoiceGuard: An Effective and Practical Approach for Detecting and Blocking Unauthorized Voice Commands to Smart Speakers has been published in DSN 2023. [PDF]

• Adversarial Machine Learning Attack against Graph-based Classification System
• Proposed a new attack by adding fake nodes to a graph without modifying existing edges to fool graph-based classification method. Formulated the proposed attack as an optimization problem and solved it using C++.
• Evaluated the effectiveness of the proposed attack on the Facebook dataset and Enron dataset.
• The paper Attacking Graph-Based Classification without Changing Existing Connections has been published in ACSAC 2020. [PDF]

• Air Force UAV Authentication Schemes
• Proposed two authentication schemes for two-way mutual authentication between the UAV and dumb sensors.
• Implemented prototype systems in Python on Raspberry Pis to evaluate the proposed schemes.
• The paper Effective UAV and Ground Sensor Authentication has been published in IEEE GLOBECOM 2019. [PDF]

Education

• Jan 2022 - Aug 2025
• Stevens Institute of Technology
• Ph.D. in Computer Engineering

• Sep 2019 - Dec 2021
• Temple University (Continued at Stevens Institute of Technology)
• Ph.D. Program in Computer and Information Sciences

• Sep 2017 - May 2019
• Temple University (Dual Bachelor's Master's Degree program)
• M.S. in Computer Science

• Sep 2014 - Jun 2018
• University of Science and Technology of China
• B.S. in Mathematics and Applied Mathematics

Work
Experiences

• May 2024 - Aug 2024
• Keysight Technologies @ Austin, TX
• Security Research Intern
• Conducted security testing on IoT devices using fuzzing techniques over Wi-Fi and Bluetooth to identify vulnerabilities.
• Conducted chip debugging via UART, SPI, and SWD interfaces using PuTTY, flashrom, OpenOCD and GDB.
• Performed firmware analysis with Ghidra and utilized QEMU along with GDB for firmware emulation and debugging.
• Revealed two critical vulnerabilities on Eufy Smart Lock and disclosed them to Eufy Security Team with detailed documention and video demo. Received an $800 bounty from Eufy for each disclosure ($1,600 in total), ranked 5th.

• Aug 2024 - Dec 2024
• Stevens Institute of Technology
• Teaching Assistant for the course ENGR 245: Circuits and Systems

• Jan 2022 - Current
• Stevens Institute of Technology
• Research Assistant

• Sep 2019 - Dec 2021
• Temple University
• Research Assistant

• Aug 2021 - Dec 2021
• CIS 3342: Server-side Web Application Development - Temple University
• Teaching Assistant & Lab Instructor

• Jan 2021 - May 2021
• CIS 4378: Computer and Network Security - Temple University
• Teaching Assistant & Lab Instructor

• Aug 2020 - Dec 2020
• CIS 4319: Computer Networks and Communications - Temple University
• Teaching Assistant & Lab Instructor

• Jan 2020 - Dec 2021
• CIS 3319: Wireless Networks and Security - Temple University
• Teaching Assistant & Lab Instructor

• Jan 2020 - May 2020
• CIS 3211: Automata, Computability, and Languages - Temple University
• Teaching Assistant