Five Steps Towards IoT Device Security

This past summer, the Trusted Computing Group (TCG) published an interesting document entitled, Architect’s Guide: IoT Security. The document outlined 5 critical strategies that developers of connected devices should consider to help gauge security risks and develop appropriate security controls to manage or reduce the risks. Given the heightened awareness of the potential vulnerabilities of connected IoT devices and networks to viruses, malware, industrial espionage, and other types of cyber threats, I thought it would be a good time to review these recommendations.

As noted in the document, the security challenges presented by the IoT are many and formidable and are being widely discussed – for example, consider the unprecedented number and variety of connected devices in the near future; a patchwork of highly heterogeneous networks involving many device manufacturers; legacy equipment and industrial control systems that often co-exist with traditional IT networks; unattended and unmanaged devices that are difficult or impossible to access for physical updating; and many other potential risk areas.

To address these challenges, the TCG outlined a 5-step process to help developers build in security from the initial concept to the final product. Following is a brief summary of these 5 steps. You can download the complete document here.

Step 1: Assess IoT Goals and Risks
Gaining a full understanding of the strategic goals of the IoT deployment is an important first step towards understanding the risks. Once goals are defined, sketch an architecture diagram and show how data and control flows through the system. Add security controls to the diagram to address the salient risks. Finally, document the risks that could threaten the system, focusing on three critical security properties: confidentiality, integrity, and availability.

Step 2: Manage Identity and Integrity
Only authorized parties should be able to gain access to the system to ensure its integrity and components are not compromised. If compromise cannot be prevented, it must be readily detectable and remedied.

Step 3: Encrypt Confidential Data
Data in transit as well as stored data should be protected with encryption. For long lived systems, plan for key updates and changes in cryptographic algorithms. Carefully consider where encryption keys will be safely stored and inaccessible to unauthorized users.

Step 4: For Critical Systems, Use Hardware Security and Standards
Software bugs can be exploited to compromise IoT systems. Critical components should be protected by security hardware as this approach helps protect against malware and attacks that are typical in vulnerable software.

Step 5: Protect Limited Devices with Overlay Networks
Many IoT systems included limited devices that cannot be upgraded to include security capabilities. These systems can be protected by placing them on an “overlay network” that insulates them from attacks and protects the confidentiality of the data traffic.

I highly recommend that IoT architects read this document in more detail as well as many other resources available on the Trusted Computing Group’s website.

Also, if you are attended RSA 2016, you might have joined us in a security session organized by the Trusted Computing Group. During the session, Wibu-Systems and Infineon Technologies presented a live demonstration on “IP Protection and Flexible Licensing Applied to TPM Connected Devices”. In addition to all the security features that safeguard the intellectual property of your applications by binding the protected license to an Infineon OPTIGA™ TPM 2.0 in the target system, CodeMeter also showed its strength on a separate front: the monetization of your software through a lean and flexible licensing system.