Guidelines for securing industrial networks

01 September 2020

Because component suppliers are playing an increasingly important role in Industrial Internet of Things (IIoT) networks, it is useful to understand the security requirements they must meet when designing devices for deployment on them. Susan Lan reports.

In 2002, the International Society for Automation (ISA) produced the ISA-99 document to advise businesses operating in automation industries how to protect against cybersecurity threats. Since publication, the ISA documents have been aligned with those more frequently used by the International Electrotechnical Commission (IEC). Currently, the IEC 62443 standard constitutes a series of standards, reports, and other relevant documentation that define procedures for implementing electronically secure Industrial Automation and Control Systems (IACS). Following these guidelines can significantly reduce the chance of a successful cyber attack.

IEC 62443 guidelines define four security threat levels. The security standard level 2 is the baseline requirement of the automation industry. It relates to cyber threats posed by hackers, which is the most common attack experienced by system integrators who secure industrial networks. Level 1 is to protect against accidental unauthenticated access and Levels 3 and 4 are against intentional access by hackers who utilise specific skills and tools.

Within the standard are several subsections that relate to different parties. The component requirements are derived from foundational requirements, including identification and authentication control, use control, data integrity and confidentiality, as well as backup for resource availability. 

Infrastructure: If a network component allows users to access devices or applications, the network component must be able to uniquely identify and authenticate all users, including humans, processes, and devices. This allows separation of duties and the principle of least privilege that ensures every user only has access to information and devices that are essential for them to be able to perform their role within the network. It is essential to avoid the unnecessary security risk of granting users greater access to the network than is necessary. Following this guideline will help secure the infrastructure of a network and provide a solid foundation to develop networks.

Account management: The ability to support the management of accounts – including establishing, activating, modifying, disabling, and removing accounts – must be supported across the network. This ensures that no accounts are created, modified, or deleted unless permission has been granted, and forbids embedded devices from making any unauthenticated connections. The management of accounts feature has several possible scenarios, which if not implemented, could cause problems for asset owners. For example, a person who works on the network gets promoted, so they now require more access to devices and applications, and their privilege level must be adjusted accordingly. Another frequently encountered example is when an employee leaves an organisation. As soon as they cease being an employee they must have their network privileges revoked. 

Identifier management: Any component of the network with a direct user interface must directly integrate into a system that identifies individuals by user, group, role, and/or system interface. This stops users from being able to access devices connected to the network that they have not been granted access to. As those with different roles on a network have different privileges, a network administrator’s account can often manage device configurations on a network, but someone who has guest level access can only view devices, but not alter configurations. In addition, there should be security procedures in place if an account has not been accessed for a certain period of time that allows the account to be deactivated. The identifier management feature controls each user’s account on the network and ensures that users are confined to just the roles assigned to them by network administrators.

Authenticator management: All devices on a network must be able to confirm the validity of any requests for system/firmware upgrades, and verify that the source is not trying to upload any viruses or malware. This is achieved by requiring the use of tokens, keys, certificates, or passwords. If no authenticator management system is in place, anyone wishing to attack the network could very easily upload malware, allowing them to change settings or take over control of the network.

Password-based authentication: For network components that utilise password-based authentication, the network component must integrate a password policy that enforces the following:

• The password composition must state what type of characters are allowed, as well as the number of characters required before a password will be accepted as valid.
• The frequency that the password must be changed.

A password is a simple way for network administrators to protect their network without requiring any additional work from system engineer. Utilising an effective password policy will keep out the majority of hackers who gain access to networks by using brute force to break weak passwords.

Public key authentication: Public key authentication should be used to build a secure connection between servers and devices, or device-to-device connections. To enable this function, each network component must be able to validate certificates by checking the authentication of the signature, as well as the revocation status of a certificate. In addition, it should construct a certification path to an accepted certification authority, or in the case of self-signed certificates deploy certificates to all hosts that communicate with the subject to which the certificate is issued. Public key authentication is important because it stops information from being sent to the wrong place, and also stops confidential information that should remain within the network from being transferred to unverifiable sources outside.

Use control: All the devices that appear on a network must support login authentication. To restrict unwanted users from gaining access to a device or the network, it is necessary to limit the number of times a user can enter the password incorrectly before being locked out. As the majority of attacks on industrial networks are performed by hackers using brute force attacks, login authentication is an effective method of stopping hackers from gaining access to a network. In addition, the system or device must also be able to inform users whether their login attempt was successful or not. Informing users that they are logged into the network allows them to confirm their current status and proceed knowing that changes or alterations they make to network settings or devices have been authenticated.

Data integrity: Data integrity plays a vital role across all IIoT networks. It ensures that data is accurate, and that it can be processed and retrieved reliably. There are several security measures that can be utilised to protect the data, including SSL, which enables encryption between a web browser and a server. As data is constantly moving around a network, network operators need to be sure that the data is moving in a safe, reliable, and efficient manner. If the data is sent to unintended recipients, the network operators will not only lose control of their data, but also leave their networks vulnerable to hackers.

Backup for resource availability: All of the applications or devices that are found on a network must be able to back up data without interfering with network operations. The main advantage of performing regular backups is to ensure that no data is lost and that if the network experiences some problems the network can utilize the data that has been backed up to return the network to normal. In addition, the backup process must ensure that any private information that is on the network is stored in accordance with data protection policies and is not accessible by anyone who should not have access to that information. In some cases this means that data can’t be stored outside the network. Any data breach containing users’ personal information is extremely damaging to network operators as well as to those whose data has been accessed by those it shouldn’t be accessed by.

As more devices are added to networks, the security of these devices is of paramount concern to asset owners. It is acknowledged throughout the industry that adopting the best practice approach to security gives asset owners the best chance of protecting their network from those with malicious intent. The complete system-level security must be built upon the foundations that consist of each individual component's security functions.

Susan Lan is a product manager at Moxa.

Contact Details and Archive...

Related Articles...

Additional Information...

Print this page | E-mail this page