Originally, "defense-in-depth" referred to a military strategy, in which several lines of defense are built one after the other - in the "depth of space". This principle is now applied to a company's cybersecurity:
With the "defense-in-depth" strategy, the security architecture is divided into different layers. Each layer contains different security measures. If an attacker breaks through a layer, for example a firewall, he will be stopped again by the next security measure.
Get every important blog post or new information Eaton is pulishing for machine and systembuilders.
Although there are differences between traditional IT systems and ICS (Industrial Control Systems), the basic concept of "defense-in-depth" is applicable to both. The usual operational security levels are listed below.
To better understand the principle of "defense-in-depth", let's look at an engineering workstation (EWS) in a control center of an industrial facility as an example:
The physical protection of components such as PLCs, storage media, workstations as well as laptops and servers is an important element of cybersecurity. An example of this is a SCADA system whose components are geographically distributed (e.g. in power grids or oil and gas infrastructure). These facilities have stations (e.g. substations) in remote locations. They should have very strong physical safeguards for protection (e.g. electric fences), and surveillance (e.g. CCTV cameras). A firewall cannot prevent an attacker from stealing hardware such as a storage device from a station.
Physical security helps defend against threats such as unauthorized entry, theft, vandalism, or even natural disasters such as fire and flood. It is implemented with physical facilities such as locks, video surveillance, protective barriers, rooms and cabinets with access control mechanisms, uninterruptible power supplies and security personnel to protect hardware and software assets.
The Internet is based on the TCP/IP protocol family. It enables a connection for hosts belonging to the network. Due to the structure of the Internet, however, it is almost impossible to control the path taken by data packets. A user can never be sure who is actually reading the data and what is being done with it.
Network security refers to technologies and processes that ensure the security of the network itself and the data transmitted in. Important aspects are confidentiality and integrity. Unauthorized intrusion into the networks should be prevented. In addition to organizational measures, both hardware and software technologies are used to prevent malicious actors from accessing or moving in a network.
Network security can begin with simple username and password authentication. However, depending on an organization's risk profile, multi-factor authentication and the usage of complex hardware and software technologies may also be necessary. Examples could be networkfilter and Intrusion Detection Systems (IDS).
Commonly important elements of network security are “Demilitarized Zones” (DMZ). This are physical or logical sub-networks in which critical components are placed. A DMZ groups systems and components that are required in both internal and external networks and secures the company network, by isolating it with firewalls. At least a three-tier architecture is recommended in industrial facilities. A DMZ separates the general corporate IT network (this is where the hosts for e-mail and web servers are located, for example) from the operational network in which the components of the ICS are located. The DMZ adds an additional layer of security to a company's operational network. Hence, an external attacker can only access the devices in the DMZ and not other parts of the network.
In principle, a host is nothing more than hardware within a network on which software can be executed. In industrial control systems, a host can be a PLC, an HMI, or a server. Host security ensures that, among other things, hardware, software, server and storage components are protected against cyberattacks, and this also includes regular updates, monitoring and access controls.
Applications are the most used and exposed interfaces of a device. This makes them a prime target for cyberattacks - and explains why application security is critical to a defense-in-depth strategy. There are various ways to protect the applications: input control, authentication and data encryption are just a few examples.
Understanding what is happening inside the network from both a performance and security perspective is critical to cybersecurity. This is especially true in a control system. Security Information and Event Management (SIEM) helps organizations monitor infrastructure components and identify potential security threats and vulnerabilities. Today, thanks to the power of AI and machine learning, SIEM offers advanced analytics of user and entity behavior.
With today's threat situation, perimeter protection (Firewall, virus and spam filters) at the edge of networks is no longer sufficient - especially not with industrial control systems (ICS). Rather, it requires a multi-layered "defense-in-depth" that protects manufacturing systems such as the various walls and ramparts of a castle. In the white paper “Cybersecurity considerations for industrial control systems”, Eaton describes the measures that users must take to protect industrial controllers and automation components from cyberattacks and how manufacturers develop "secure" products.
Contact our team or sign up to ask questions and stay up-to-date on news, product updates and industry trends.