Editor's Choice


Industrial control system cybersecurity - Part 5: ICS network segmentation.

October 2018 Editor's Choice IT in Manufacturing

In the last three articles on cybersecurity in ICS environments, we have covered risk assessments, asset discovery and vulnerability management, environment hardening and security monitoring. In the penultimate article, we will cover network segmentation in ICS networks.

Historically, many ICS/engineering departments were not focused on protecting the inside of their networks, only the perimeter was protected with the firewall being seen as the single line of defence against the malicious insiders, third-party vendors and the bad guys from the outside. This strategy, while effective for its day, does not hold true in the modern digital world. Today’s attacks are being facilitated by large and well-funded groups of cyber criminals looking to steal intellectual information, stop production and extort companies. Once access is gained by breaching the perimeter, these cyber criminals are able to move freely within your network. This is why it is strongly recommended to implement a network segmentation framework.

Splitting up the network

ICS network segmentation is the process of splitting up your network into different segments or sub-networks, to improve performance, but more importantly, to make it more difficult for an adversary to freely move around if they compromise a part of your network. To define this further, it is the process of grouping similar assets and then enforcing a segment between the levels both above and below.

To put this into perspective, Target Corporation, a leading USA retailer, lost 40 million credit and debit card numbers in December 2013. The first part of this compromise is that the cyber criminals stole credential information from a third party HVAC supplier. The second part is that these credentials were then used to gain access to the Target Corporation network. The third part is that once the cyber criminals gained access they targeted the POS systems, by installing malware on them. There is more to this incident (an entire article on its own), but it does highlight the need for strong effective network segmentation. If there was proper network segmentation between the POS network, the third party network and the main corporate network, it would have been much more difficult to steal the information.

Purdue Enterprise Reference Architecture

One of the most commonly used models is that of the Purdue Enterprise Reference Architecture model, more commonly known as PERA or just the Purdue model. I strongly urge all of those responsible for ICS cybersecurity to review this method. It was developed by the Industry-Purdue University Consortium for Computer Integrated Manufacturing, and has been widely adopted by major industrial control system cybersecurity frameworks such as NIST 800-82 and ISA/IEC 62443.

From a hierarchical view the model is comprised of 6 levels and 5 zones. The 6 levels are:

• Level 0: Process.

• Level 1: Basic control.

• Level 3: Operations and control.

• Level 4: Business planning and logistics.

• Level 5: Enterprise network.

And the five zones being:

• Enterprise zone.

• Demilitarised zone (DMZ).

• Manufacturing zone.

• Cell/area zone.

• Safety zone/Safety Instrumented System (SIS).

The diagram is a very basic control network depicting how the Purdue model should logically be implemented.

One aspect to take note of from the diagram is that no control system protocol should traverse the ICS network into the enterprise or business network. All too often we still find ICS traffic on the IT network(s), which not only slows down network performance by having unnecessary traffic ‘on the wire’, but also provides huge security risks as these protocols have no, or very limited, built-in security. If ICS traffic is absolutely required to traverse the ICS network through to the IT network, ensure that is it is strictly controlled.

Each ICS system is different and requires certain tweaks and changes to the customer’s specific ICS network segmentation framework. Where the Purdue model helps is that it assists in designing a base framework which you can then build on. As I’ve stated previously, there is no ‘one size fits all’ framework that is right for everyone, and there are other models that you might want to consider to suite your organisation’s needs. The Industrial Internet of Things (IIoT) and Software-Defined Networking (SDN) is also changing the way we see and segment our networks.

Tommy Thompson

Tommy Thompson is a passionate cybersecurity professional with some 15 years’ experience. Starting as a firewall engineer in 2001, Thompson has assisted a variety of companies in numerous roles with their cybersecurity problems. He holds a BComm degree in Information Management from Oxford Brookes University (UK) and he is certified by PECB (Canada), as a Scada Security Professional (CSSP).

For further information contact Tommy Thompson, +27 (0)11 463 0096, tommy@nclose.com





Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

How to size and select a servo motor
Festo South Africa Editor's Choice Motion Control & Drives
Festo highlights some factors to consider in the process of sizing and selecting a servo motor effectively to ensure optimal performance, reliability and energy efficiency.

Read more...
Dynamic control of industrial solar plants and energy storage systems
Beckhoff Automation Editor's Choice Electrical Power & Protection
Spanish Group, Power Electronics has demonstrated its comprehensive expertise in sustainable energy supply in over 3000 solar and energy storage projects with a total installed capacity of 120 GW. To control its modular systems, the company relies on open, high-performance Beckhoff control technology.

Read more...
Loop signature Part 2-4: Feedforward Control: Part 3
Michael Brown Control Engineering Editor's Choice Fieldbus & Industrial Networking
In the previous articles in this series, the basic theory behind feedforward control was discussed, and it was also shown how to apply feedforward in practice. In this article, it will be shown how well feedforward can work in practice by giving a couple of examples.

Read more...
Engineers, AI and the risk of cognitive surrender
Editor's Choice
AI will have a significant impact on how engineering work gets done in the future, but the cognitive implications of continued AI use are serious.

Read more...
Reinventing grain silo management
VEGA Controls SA Editor's Choice
The VEGAPULS 6X radar sensor is designed for continuous level measurement to help overcome the challenges faced by storage in grain silos.

Read more...
Trends in humanoid robots
Editor's Choice
Humanoid robots are increasingly viewed less as futuristic prototypes and more as a practical route to bring artificial intelligence into human-designed environments.

Read more...
Four futures for AI: The choices we need to make now
Editor's Choice IT in Manufacturing
AI is everywhere and its implications are now structural. The question is no longer whether AI will matter, but what kind of society it will shape.

Read more...
Modular control platform for the hydrogen industry
Beckhoff Automation Editor's Choice Electrical Power & Protection
With a seamless modular control solution from Beckhoff featuring over 500 data points and numerous ELX series terminals with intrinsically safe interfaces, Greenlight Innovation is breaking new ground in hydrogen testing.

Read more...
Loop signature Part 2-3: Feedforward Control: Part 2
Michael Brown Control Engineering Editor's Choice Fieldbus & Industrial Networking
Feedforward control tuning is not nearly as critical as feedback tuning, and fairly simple models are usually fine for the purpose in hand.

Read more...
Proactive treatment of industrial boiler water
Editor's Choice
As water treatment is a critical aspect of industrial boiler management and potentially one of the greatest operational risk points, AES relies on close partnerships with third-party industrial water treatment specialists. These act as important safety nets.

Read more...









While every effort has been made to ensure the accuracy of the information contained herein, the publisher and its agents cannot be held responsible for any errors contained, or any loss incurred as a result. Articles published do not necessarily reflect the views of the publishers. The editor reserves the right to alter or cut copy. Articles submitted are deemed to have been cleared for publication. Advertisements and company contact details are published as provided by the advertiser. Technews Publishing (Pty) Ltd cannot be held responsible for the accuracy or veracity of supplied material.




© Technews Publishing (Pty) Ltd | All Rights Reserved