SGER: Architecting Effective Computer Security Grand Challenge Competitions

提案编号：0749648PI: Nicholas Weaver机构：加州大学伯克利分校国际计算机科学研究所标题：SGER：构建有效的计算机安全大挑战竞赛提案摘要：一个大挑战竞赛（GCC），如DARPA机器人大挑战或安萨里X奖可以激励社区，即使是那些目前没有资助的，产生可以产生变革性影响的研究成果。有人建议NSF在计算机安全领域建立一个GCC。与NSF和其他机构合作，PI确定了一些伴随这样一个计算机安全竞赛的困难，包括需要确保社区的利益，发展一个公平和可信的规则集来评判参赛作品，以确保任何竞赛既呈现出有效的公众形象，又将产生有用的结果。基于这些见解，PI正在制定规则，与几所大学合作，或在研讨会或会议上举行初步和小规模的竞赛。如果成功，完整的GCC将被大规模设计和提供。一个关键问题是竞争的例子。PI正在设计基于自我防御网络的比赛，其中竞争对手将获得一个参考系统，他们的自动化工具（防御）必须尝试在没有人为干预的情况下发现和删除漏洞。

Proposal Number: 0749648PI: Nicholas Weaver Institution: International Computer Science Institute, University of California Berkeley Title: SGER: Architecting Effective Computer Security Grand Challenge CompetitionsProposal summary: A Grand Challenge Competition (GCC), such as the DARPA robot grand challenge or the Ansari X Prize can energize the community, even those not currently funded, to produce research results that can have a transformative impact. It has been suggested that NSF create a GCC in the area of computer security. Working with NSF and other agencies, the PI identifies some of the difficulties attendant to such a competition on computer security, including the need to ensure the interest of the community, to develop a fair and credible rule set for judging entries, to ensure that any competition both presents an effective public face and will produce useful results.Based on these insights, the PI is developing rules for an initial and scaled-down competition to be held in conjunction with a few universities or at a workshop or conference. If successful, the full GCC would be designed and offered on a large scale. A key issue is the competition example. The PI is designing the competition based on a self-defending network, wherein the competitors would be given a reference system which their automated tools (the defense) must attempt to discover and remove vulnerabilities without human intervention.