DAWN SONG
Assistant Professor, Computer Science and Electrical and Computer Engineering
www

My primary research focus is in computer security and cryptography. I have tackled security problems in a wide spectrum of areas, including operating systems, networking, sensor networks, databases, software engineering, and electronic commerce. My main research goal is to develop novel techniques and fundamental principles for building secure and catastrophe-resilient systems that can serve as the critical infrastructure for our society. I am working on the following projects:

Internet Security Internet is becoming part of our critical infrastructure as the our society, economy, and government are increasingly relying on the Internet. However, the Internet is extremely vulnerable -- for example, large scale Distributed Denial-of-Service (DDoS) attacks and Internet worms disrupt critical Internet services and cause significant financial loss and operational instability. This project focuses on designing new techniques and systems to automatically detect Internet attacks, identifying origins of attacks, and throttle and mitigate attacks. These techniques need to adapt and react to traffic change on the Internet fast and automatically and need to be backwards compatible with exising systems and support incremental deployment.

Automatic Tools for Building Secure Systems Building a secure system is one of the most complex and error-prone processes in computing. In order to overcome the extreme complexity and subtlety of large systems, we need to create automatic tools to facilitate system builders in understanding, analyzing, optimizing, and implementing secure systems. This project aims to combine techniques from programming analysis, software engineering, model checking, and theorem proving, to build automatic and easy-to-use tools to improve the security of large and complex software systems.

Secure Pervasive Computing Pervasive computing will be important in the future. Small smart devices will be ubiquitous, providing information, services, and convenience. These devices will be low cost, and thus will have limited capabilities (low computation power, memory, storage, and battery power), and are not tamper-proof. This new environment fundamentally changes the way we gather and disseminate information and poses new challenges for security. This project focuses on designing new algorithms and cryptographic protocols in this environment, and building pilot systems to demonstrate that we can build not only a pervasive computing environment, but also a secure one.