Cybersecurity for Industry 4.0
Cybersecurity for Industry 4.0
复制标题
工业 4.0 的网络安全
DOI：
10.1007/978-3-319-50660-9
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
D. Schaefer
中科院分区：
文献类型：
--
作者：
L. Thames;D. Schaefer
传统的网络安全架构包含安全机制，提供机密性、真实性、完整性、访问控制和不可否认性等服务。这些机制被广泛用于防止计算机和网络入侵和攻击。例如，访问控制服务防止对诸如计算机、网络和数据的网络资源的未经授权的访问。然而，现代互联网安全格局的特征是攻击是大量的，不断演变的，极快的，持久的，和高度复杂的Schnackenberg等人。（2000），Anuar et al.（2010年）。这些特点给预防性安全服务带来了重大挑战。因此，能够自主检测和响应网络攻击的方法应与预防技术协同使用，以实现有效的纵深防御战略和强大的网络安全系统。这对于属于工业4.0系统的关键系统来说尤其如此。在本章中，我们将介绍如何将网络攻击检测和响应机制集成到我们的软件定义云制造架构中。本章中描述的网络攻击检测算法基于集成智能与神经网络，其输出被馈送到神经进化的神经网络预言机。Oracle生成优化的分类输出，用于为我们软件定义的云制造系统中的主动攻击响应机制提供反馈。本章的基本目标是展示如何使用计算智能方法来保护关键的工业4.0系统以及其他互联网驱动的系统。
Traditional cybersecurity architectures incorporate security mechanisms that provide services such as confidentiality, authenticity, integrity, access control, and non-repudiation. These mechanisms are used extensively to prevent computer and network intrusions and attacks. For instance, access control services prevent unauthorized access to cyber resources such as computers, networks, and data. However, the modern Internet security landscape is characterized by attacks that are voluminous, constantly evolving, extremely fast, persistent, and highly sophisticated Schnackenberg et al. (2000), Anuar et al. (2010). These characteristics impose significant challenges on preventive security services. Consequently, methodologies that enable autonomic detection and response to cyberattacks should be employed synergistically with prevention techniques in order to achieve effective defense-in-depth strategies and robust cybersecurity systems. This is especially true for the critical systems belonging to Industry 4.0 systems. In this chapter, we describe how we have integrated cyberattack detection and response mechanisms into our Software-Defined Cloud Manufacturing architecture. The cyberattack detection algorithm described in this chapter is based on ensemble intelligence with neural networks whose outputs are fed into a neuro-evolved neural network oracle. The oracle produces an optimized classification output that is used to provide feedback to active attack response mechanisms within our software-defined cloud manufacturing system. The underlying goal of this chapter is to show how computational intelligence approaches can be used to defend critical Industry 4.0 systems as well as other Internet-driven systems.