Research Interests

My research interests lie widely in the area of wireless, mobile, embedded systems and security. I focus my eyes on security and privacy threats in Internet of Things (IoT) and Cyber Physical Systems (CPS). Specifically, my research utilizes physical layer attacks to achieve application layer threats. I also design and implement attack and defense methods on networking systems (such as LTE, GPS) using techniques like Software Defined Radio (SDR) and Deep Learning.


  • GPS Stealthy Manipulation: To divert mobiles to malicious locations, naive GPS spoofing attacks may never be practical. Therefore, we developed spoofing to stealthy manipulation. the goal of it is to use GPS spoofing techniques to trigger turn-by-turn navigation to guide the victim to a wrong destination without being noticed. Our key idea is to slightly shift the GPS location so that the fake navigation route matches the shape of the actual roads. By that means, we can trigger physically possible instructions that leads to dangerous places. We designed iterative attacking algorithm and measured the effects in physical enviornment and on actual users. This project is made possible by the support of Secure Localization Team.

  • MAPRO & SecRF: GNSS spoofing has huge threats on widely-adopted devices such navigators and base stations. We introduce MAPRO as a sheild and filter to malicious GNSS spoofing signals. It’s an embedded device which can be easily carried and settled beside valuable infrastructures. The key idea is to use SDR platforms to tranceive and process on GNSS signals(GPS, Beidou, GLONASS). Features are extracted and signals are categorized by a CNN model we trained. Then we use a technique called friendly spoofing to adjust physical signals to a normal state. We extend the frequency band to get more protections on other RF applications and then comes SecRF. MAPRO & SecRF are products we designed to secure GPS/RF-assisted devices in Dominity Security Technology Co., Ltd.

  • DeepDroneIndicator: Drones appearing in airport and private places are threatening, because those “flying laptops” disorder the aero-security and have capacity to get privacy without being noticed. Therefore, DeepDroneIndicator is a project focusing on detecting the flying directions of a drone. By monitoring the control loop(wireless channels) of a drone, its appearance and motions can get indicated. Further implementation involves localizing the drone and intelligently landing it by smart GPS and ADS-B spoofing. This is a project during my 2017 summer internship at 360 Unicorn Team.

Technical Skills

  • Software Defined Radio: GNU Radio(HackRF, LimeSDR, USRP)
  • Deep Learning: Tensorflow(CNN, LSTM, etc.)
  • Programming Language: C, C++, Python, Java
  • Development: Web(LAMP and Rails stack), Android
  • Embedded Platforms: Raspberry Pi, Arduino