Understanding the dynamics involved in industrial IoT attacks is crucial to perform security risk analysis and mitigation. Threat modeling is commonly used as a security countermeasure, and has been discussed later in this chapter. Attack and fault trees are two methodologies useful to develop security threat models and to communicate the risk of an attack.

In the real world, most attacks are highly customized to target specific vulnerabilities in IoT products and connectivity. Many attacks target zero-day vulnerabilities. In the case of zero-day vulnerabilities, an exploit already exists and can be easily proliferated through the internet or corporate networks to create a snowball effect. Since IIoT involves significant investment and skills, most attacks involve nation state threat actors, who are motivated to create a major impact...

As already noted in this book, the concept of securing cyber-physical systems is a superset of what we normally understand by cybersecurity and information security.

To properly represent the scope of IIoT security, the term trustworthiness is used (NIST-CPS) (IIC-IISF). A working definition of trustworthiness for CPS, according to NIST-CPS, is:

Trustworthiness of an IIoT system is an important stakeholder expectation. To make an IIoT system trustworthy, security characteristics of both IT and OT domains must be combined (IIC-IISF). As shown in Figure 2.6, the key characteristics of a trustworthy...

Data is the prime asset in the IIoT value chain. Industrial devices such as sensors, actuators, and controllers generate state and operational data. The information inherent in this industrial big data enables a variety of descriptive, prescriptive, and predictive applications and business insights. This end-to-end flow of data, from the point of ingestion, through information processing using various extract, transform and load (ETL) functions, applying AI and machine learning intelligence, up to the point of data visualization and business application, is collectively referred to as the industrial big data pipeline (shown in Figure 2.7):

The preceding diagram is explained as follows:

• On-premise data sources: On-premise data includes usage and activity data...

In 2015, the IIC released the Industrial Internet Reference Architecture (IIRA) for IIoT systems (IIC-IIRA). It uses "ISO/IEC/IEEE 42010:2011 Systems and Software Engineering–Architecture Description" for architectural conventions and common practices. IIRA provides an architectural framework to analyze concerns, views, models, and so on with certain degrees of abstraction. The use of reference architectures helps to incorporate security by design. Architects can build use case-specific IIoT architectures on top of these reference architectures.

In this section, the four viewpoints of IIC's reference architecture are briefly discussed. These viewpoints simplify the understanding and decomposition of IIoT architectures. You can find an in-depth treatment of these viewpoints in (IIC-IIRA).

This chapter presented a primer of attacks, countermeasures, and threat modeling, which lays the foundation for effective risk analysis and mitigation. It also provided the readers with insights into the distinguishing characteristics of trustworthiness for IIoT systems and the functional components of the industrial big data pipeline.

IIoT systems are highly complex; this chapter presented IIoT architectural viewpoints and patterns as developed by the IIC to provide you with a crisp understanding of end-to-end IIoT system components. Based on usage, operations, and functional domains, the IIoT security architecture was decomposed into a four-tier security model, which has been further elaborated in the subsequent chapters of the book.