A security professional considers the small memory sizes of Internet of Things (IoT) systems as physical limitations to security, privacy, communication, computation, and trusted sensing. Security challenges of designing new IoT systems with sensors and actuators have been addressed by CAD designers. Rising semantic attacks on data being processed by IoT systems create an opportunity for redesigning IoT systems to better secure the data for less energy.

The authors take up these challenges by addressing security requirements of IoT systems, such as authentication, privacy, and data revocation (data forgetting), while ensuring compliance with standard Internet protocols and services. They propose several hardware-based IoT security approaches as more effective ways to meet security challenges than the software-based approaches. They conclude that the use of stable physical unclonable functions (PUFs) and digital public physical unclonable functions (PPUFs) for several IoT security protocols would best fit the need for tightening hardware security requirements while reducing energy consumption.

The authors keep the presentation informal with illustrations to show the differences between proposed hardware approaches. For those interested in the challenges and opportunities of designing security for IoT systems, this paper is definitely worth reading.