Department of Computer Science,
University of California, Irvine
Office: ICS1 420 (inside ICS1 418), Inner Ring Rd
Email: alfchen at uci dot edu
The 4th AutoSec (Automotive & Autonomous Vehicle Security) Workshop
has happened with
NDSS on April 24, 2022 (hybrid)!
[Final Program & Awards] [CFP] Follow @autosec_conf Tweet to @autosec_conf
Join us on April 24 (hybrid) for 2 in-person keynotes (academia+industry), 15 papers + 15 demos on latest security/privacy research on automotive, autonomous vehicles & vehicle networks, awards annoucement, and new sessions for community shout-out & future planning!
Come here for AI security research in autonomous driving (AD) & smart transportation?
Collection of our latest efforts is at our AD & CV Security Website and YouTube channel
» (2022) [arXiv'22] Systematization of Knowledge (SoK) of the semantic AD AI security research space + Proposal of PASS platform «
» (2022) [NDSS'22] PlanFuzz: First automated method to discover security vulnerabilities specific to AD planning (semantic DoS-focused) «
» (2022) [Workshop] AutoSec (Automotive & Autonomous Vehicle Security) @NDSS'22 (15 papers+15 demos) @autosec_conf «
» (2021) [Usenix Sec'21] DRP attack: Seemingly-benign dirty road pattern can attack self-driving & cause lane departure in 1 sec «
» (2021) [Contest] AutoDriving CTF'21 @ DEFCON (one of world's largest & most notable hacker conventions) Follow @autodrivingctf «
» (2021) [IEEE S&P'21] MSF-ADV attack: Fundamentally challenge sensor fusion based AD perception in practical settings «
» (2021) [Workshop] AutoSec (Automotive & Autonomous Vehicle Security) @NDSS'21 (17 papers+10 demos) «
» (2020) [Usenix Sec'20] FusionRipper: First security analysis of fusion based AD localization, discover strategic GPS spoofing to break fusion «
» (2020) [Usenix Sec'20] First general adversarial sensor attack & defense for LiDAR-based perception in autonomous driving «
» (2019) [ACM CCS'19] First security analysis & adversarial sensor attack design for LiDAR-based perception in autonomous driving «
» (2018) [NDSS'18] First security analysis & congestion attack design for CV/V2X-based intelligent traffic light control «
Research positions available for self-motivated students interested in computer security, especially AI/software/network security in exciting emerging CPS/IoT systems such as self-driving cars (link1, link2), intelligent transportation (link1, link2), and smart home (link1, link2).
BS/MS students: Please fill out this form if looking for research experience in my group.
I am teaching the graduate-level systems security course at Spring'22: CS 205: Computer and Systems Security
[News] (12/2021) [Award] Thrilled to annouce that I just received the NSF CAREER award! Excited to leverage this award to keep exploring AI security in the emerging Internet of Autonomous Things (IoAT) such as autonomous cars, drones, and robots.
[News] (05/2021) [Award] Congratulations to Kanglan Tang, an undergraduate student in my group, for receiving the prestigous UCI Chancellor’s Award for Excellence in Undergraduate Research! This award is given to only 1 student per school. Kanglan is extremely excellent with a 1st-author paper published, presented herself, and accepted for demos at multiple academic venues, and also awarded multiple travel grants to top-tier security venues. Look forward to her continued success at UC Berkeley!
[News] (04/2021) [Award]
undergrad cyberseucrity team Cyber@UCI
finished 5st place nation-wide at
National CCDC, out of 168+
university/college teams in US!
Absolutely a huge achievement for the first time at National. Can't wait for
next year :)
>> News coverage: UCI Finishes in Top 5 at National Collegiate Cyber Defense Competition
[News] (04/2021) [Award] Our
undergrad cyberseucrity team Cyber@UCI won
the 1st place (Gold Medal) at
CCDC Western Regional, and advanced to
National CCDC for the first time!
We beat many strong rivals such as Stanford, UCSD, ASU, and UCR. Super proud of
>> News coverage: UCI advances to national cyber defense competition following “historic” regionals win
[News] (09/2020) [Award] My
research group ASGuard (Autonomous Systems Guard) won the
(1st place) at
first Autonomous Driving Security CTF organized by Baidu Security! Was quite
a close game, and we were able to beat 20+ teams! Congratulations to my students Junjie Shen (Captain), Ningfei
Wang, Takami Sato, Ziwen Wan, Yunpeng Luo, and BS Alumni Zeyuan Chen (now at
>> News coverage: ICS Team Takes 1st Place at Baidu’s AutoDriving Capture the Flag Competition
[News] (06/2020) [Usenix Security'20 (3)] Our paper on the first security analysis of Multi-Sensor Fusion (MSF) based localization in autonomous driving got into Usenix Security'20! Discovered a novel & general attack, FusionRipper, that can fundamenally defeat MSF design principle! Attack demos are in our project website. Earlier version of this work also won NDSS'19 Distinguished Poster Presentation Award.
[News] (06/2020) [Usenix Security'20 (2)] Our paper on the first general adversarial sensor attack & defense for LiDAR-based perception in autonomous driving got into Usenix Security'20! Discovered a general black-box attack based on LiDAR spoofing, and designed novel physical invariants based defense at both system & DNN model levels! More details are in our project website.
[News] (02/2020) [Award] Our poster "Security of Deep Learning based Lane Keeping Assistance System under Physical-World Adversarial Attack" wins the NDSS'20 Best Technical Poster Award (top 1/30)! Congratulations to my students Takami, Junjie, and Ningfei! Junjie has been winning NDSS Best Poster Award for 2 consecutive years, bravo!
[News] (02/2020) [EuroS&P'20] Our paper on the first study of publish-subscribe overprivilege vulnerability in autonomous driving software got into EuroS&P'20! It's the first to discover this new class of domain-specific vulnerability, and contributes not only a novel static detection method but also an efficient & module-transparent defense solution! Website & attack demos coming soon.
[News] (01/2020) [NDSS'20] Our paper on reverse engineering CAN bus commands using mobile apps got into NDSS'20! It's the first cost-effective (no real car needed) & automatic (no human intervention required) approach for reverse engineering CAN bus commands!
[News] (12/2019) [ICLR'20] Our paper on the first adversarial physical-world attack against object tracking in autonomous driving got into ICLR'20 (top-tier conference in machine learning)!
[News] (12/2019) [ICSE'20] Our paper on the first comprehensive study of autonomous driving software bugs got into ICSE'20 (top-tier conference in software engineering)!
[News] (11/2019) [Usenix Security'20 (1)] Our paper on the first security analysis of car IoT dongles got into Usenix Security'20! Identified 5 types of security vulnerabilities that can cause safety threat and car theft, with 4 CVEs newly generated (and more to come)!
[News] (11/2019) [News Coverage] Our ACM CCS'19 work on attacking LiDAR-based perception in self-driving cars is recognized by Forbes: Five Components Of Autonomous Car Security
[News] (07/2019) Please check out the project website and attack demo for our new ACM CCS'19 work on autonomous driving systems security! A collection of our recent research progress in this space is here: Connected and Autonomous Vehicle Systems Security.
I am an Assistant Professor in the Department of Computer Science at the University of California, Irvine. Before coming to UCI, I received my Ph.D. degree from Computer Science and Engineering at the University of Michigan, Ann Arbor in 2018.
My research generally focuses on network and systems security. More specifically, my research interest includes software security, network protocol security, DNS system security, smartphone system security, CPS/IoT system security, access control system security, vulnerability discovery and analysis, side channel attack and defense. Most recently, my research focuses mainly on security problems in smart systems and IoT, e.g., smart home systems, smart transportation systems, and autonomous vehicle systems.
The major theme of my research is to proactively address security challenges through systematic problem analysis and design, leveraging techniques such as static/dynamic program analysis, software testing, and network measurement. My research has developed such approaches to systematically discover, analyze, detect, and fix vulnerabilities in a wide range of important computer systems and components such as smartphone OSes, network protocols, DNS, GUI systems, access control systems, and very recently intelligent traffic signal control systems.
Selected Media Coverage (more)
My attack videos on YouTube attracted over 60,000 views (as of 09/07/2014) from all over the world, with a daily peak of over 17,000 views.
I am a Distinguished Active member of Tau Beta Pi Michigan Gamma (MI-G) Chapter.
I initialized the first Flyer Book for the CS dept. of Nanjing University, guiding students to apply for graduate schools outside China.