GRK 1076: Trustworthy Software Systems - Construction, Certification, Application

随着我们对越来越多由软件控制的技术系统的依赖，软件越来越多地影响着我们的日常生活。此外，基于互联网的应用程序无处不在，增加了我们对这些软件系统可用性的依赖。例如，汽车领域中的复杂嵌入式软件控制系统，或者电子政务和电子健康的IT系统。幸运的是，软件行业的崛起为受过学术培训的专业人士创造了就业机会，并创造了越来越多的国家价值创造。然而，这种对软件系统的依赖加剧了软件故障的后果。因此，软件系统的成功部署取决于我们对这些系统的信任程度。这种信任的相关性正在工业界得到越来越多的认识。几家软件供应商联盟计划开发所谓的可信计算平台。目前的这些举措主要侧重于安全，而信任是一个更广泛的概念。事实上，信任是由安全性、正确性、可靠性、可用性、隐私性、性能和证书等属性赋予的。因此，研究生院将通过将Oldenburg计算科学的能力与计算机法律的能力捆绑在一起，开发新的跨学科科学方法来创建可信软件系统，并培养高素质的毕业生，将构建和认证可信软件系统所需的技能转移到实践中，从而为可信软件系统的全面发展做出贡献。从技术角度来看，研究培训小组的研究方案建立并推进了基于组件的软件工程的范例。研究训练组将制定的科学方法因所调查的信任方面而异。例如，正确性是通过数学证明来证明的，而可量化的质量属性，如可用性、可靠性和性能，则需要分析预测模型，而分析预测模型需要额外的实证研究来进行校准和验证。一般来说，软件工程方法的好处必须通过案例研究和对照实验进行经验证明。

Software increasingly influences our daily life, as we depend on a raising number of technical systems controlled by software. Additionally, the ubiquity of Internet-based applications increases our dependency on the availability of those software systems. Exemplarily consider complex embedded software control systems in the automotive domain, or IT systems for eGovernment and eHealth. Fortunately, the rise of the software industry creates jobs for academically trained professionals and generates an increasing proportion of the national creation of value. However, this increased dependency on software systems intensifies the consequences of software failures. Therefore, the successful deployment of software systems depends on the extent we can trust these systems. This relevance of trust is gaining awareness in industry. Several software vendor consortia plan to develop so-called Trusted Computing platforms. These current initiatives primarily focus on security, while trust is a much broader concept. In fact, trust is given by several properties, such as safety, correctness, reliability, availability, privacy, performance, and certification.Therefore, the graduate school will contribute to this comprehensive view on trusted software systems by bundling the Oldenburg computing science competences with those of computer law to develop new interdisciplinary scientific methods for creating trustworthy software systems and to educate highly qualified graduates who transfer into practice the required skills for constructing and certifying trustworthy software systems. From a technical point of view, the research programme of the Research Training Group builds on and advances the paradigm of component-based software engineering. The scientific methods to be developed in the Research Training Group vary according to the aspects of trust under investigation. For example, correctness is demonstrated by mathematical proofs while quantifiable quality properties, such as availability, reliability, and performance require analytical prediction models, which need additional empirical studies for calibration and validation. Generally, benefits of software engineering methods must be demonstrated empirically by case studies and controlled experiments.