Projects

Project 1: High-Performance Computing

Graph is an important data representation used in many applications, such as control and data flow graph in program analysis, friendship graph in the social network, and authentication graph in APT campaigns. There are two major challenges for computing on graphs. First, Graph, in reality, is big data in terms of both graph size and count. The graph size in many real applications, e.g., social network, has reached billion scale. Second, many interesting graph algorithms have high time complexity, e.g., subgraph isomorphism is NP-complete. As real applications usually require real-time computation, it is critical to improve the scalability of graph analytics.

This project aims to design scalable graph analytics techniques and systems to both speed up the computation and scale up the accommodated graph size and count. A recent focus is on graph AI.

Publications: PeeK [SC '23], Tango [SC '23], TLPGNN [HPDC '22], Aquila [HPDC '20], iSpan [SC '18]

Fundings: NSF SHF 2331301, NSF SHF 2409211

Project 2: Cybersecurity

The computing devices, e.g., IoT devices and smartphones, are growing rapidly to reach tens of billions in scale. As the running operating systems and software are far from flawless, the vulnerabilities existent in them inevitably enable unknown attack vectors. More seriously, as libraries and code segments are often heavily reused in the software development phase, a vulnerability found in one commonly used code repository could be harnessed to compromise a large number of computing devices that are dependent on it. Therefore, identifying the vulnerabilities hidden in the devices becomes a top priority task.

This project aims to greatly improve the security of the computing devices by (1) investigating the potential attack surfaces that can cause serious damage; (2) enhancing the fundamental techniques used in existing vulnerability detection methods.

Publications: HermesSim [USENIX Security '24], API2Vec [ISSTA '23], Illuminati [EuroS&P '22], DEFInit [USENIX Security '21], BugGraph [AsiaCCS '21], APT detection [RAID '20]

Fundings: NSF OAC 2319975, NSF OAC 2419843