For instance, IoT vendors required to certify their IoT products for safety infrequently update software on these products. Users, on the other hand, may wish to update the software more often, e.g. to gain functionalities. Reports have shown how such tensions have already produced unfortunate situations where users resort to pirated IoT software (!) which only complicates matters. 47 48 PART II _ Fields of Research for IoT Cryptology 2.4 for Low-end IoT Cryptography provides the fundamental protocols and basic algorithms (primitives) for authentication, identification, and encryption on which all secure systems are built.

Cryptographers have decades of experience in the design and analysis of efficient cryptosystems and protocols for relatively powerful devices (such as PCs, servers or smartphones) on one hand, and for more constrained devices such as smart cards on the other. The rise of IoT, with ubiquitous interconnected low-power devices, brings a fascinating new challenge for cryptographers, as it mixes the application requirements of the PC paradigm with the hard physical constraints of low-end devices. Put simply, we know how to provide some security for microcontrollers on smart cards, but smart cards were never meant to be connected to the Internet; and we know how to provide Internet security for powerful processors, but not on a stringent low-energy budget. The challenge for cryptographers is to develop full-strength primitives operating within the special constraints and requirements of the IoT paradigm.

Cryptographic Primitives for Secure IoT Communications High-performance, high-security cryptographic primitives are now standardized, and widely deployed in protocol suites such as TLS (Transport Layer Security) for secure Internet communication. However, these algorithms have traditionally been developed and optimized for higher-powered platforms: from servers and PCs down to smartphones. When we move to more limited low-end IoT devices, the resource constraints are tightened to the point where conventional primitives are often too costly for the device in question. A critical challenge is thus the development, optimisation and adoption of alternative cryptographic primitives providing adequate building blocks for secure, low-power IoT communications.

Symmetric & Asymmetric Cryptographic Primitives Cryptographic primitives are divided into two fundamental classes, symmetric and asymmetric, according to the function and application of the primitive. Data encryption and data authentication, for example, are symmetric primitives; 2.4_Cryptology for Low-end IoT key exchange and signatures are asymmetric primitives. In general, symmetric primitives have much higher throughput and lower resource consumption. On the other hand, asymmetric primitives offer essential functionalities (such as digital signatures) that symmetric cryptography is literally incapable of providing. But these asymmetric primitives come at the inevitable cost of comparatively bigger keys, larger internal states, and more intensive computations, with the time, memory, and battery requirements that they entail. Optimized symmetric and asymmetric cryptography are both needed for secure IoT communication in practice.