CT-T: Collaborative Research: Complex, High-level, Integrated Properties for Security

在过去的十年中，我们在自动分析安全属性的源代码方面取得了实质性的进展。这些进步正在转化为实践，并对软件安全产生广泛的影响。然而，由于存在许多我们不知道如何分析的安全属性，这项工作的成功受到了部分限制。这个项目试图采取几个最初的步骤来扩展可以在大型软件系统上检查的安全属性的范围和范围。这项研究涉及三个方面。首先，我们将研究编程语言的设计如何支持安全编程实践。其次，我们将学习分析内存安全性的技术，这是一个复杂但基本的安全属性，令人惊讶的是，今天仍然不能静态检查。第三，我们将研究如何检查网络软件是否与标准委员会对网络协议的描述中隐含的规范一致。集成在一起，这三个方向将允许我们提供从低到高级别的安全属性的端到端保证。这个项目的目标是开始开发用于改进软件安全性的下一代技术的工作。

The last ten years have seen substantial progress in our ability to automatically analyze source code for security properties. These advances are transitioning into practice and making a broad impact on software security. However, the success of this work has been partially limited by the fact that there are many security properties that we do not know how to analyze.This project seeks to take several initial steps towards extending the range and scope of security properties that can be checked on large software systems. The research involves three tracks. First, we will investigate how the design of programming languages can support secure programming practices. Second, we will study techniques for analyzing memory safety, a a complex yet fundamental security property that, surprisingly, is still not statically checkable today. Third, we will study how to check that networking software is consistent with the specifications implied in standards committees' descriptions of network protocols. Integrated together, these three directions will allow us to provide end-to-end guarantees of security properties that span from low to high level. The goal of this project is to begin the work of developing a future generation of techniques for improving software security.