Performance-efficient cryptographic primitives in constrained devices

Abstract

Resource-constrained devices are small, low-cost, usually fixed function and very limitedresource devices. They are constrained in terms of memory, computational capabilities, communication bandwidth and power. In the last decade, we have seen widespread use of these devices in health care, smart homes and cities, sensor networks, wearables, automotive systems, and other fields. Consequently, there has been an increase in the research activities in the security of these devices, especially in how to design and implement cryptography that meets the devices’ extreme resource constraints. Cryptographic primitives are low-level cryptographic algorithms used to construct security protocols that provide security, authenticity, and integrity of the messages. The building blocks of the primitives, which are built heavily on mathematical theories, are computationally complex and demands considerable computing resources. As a result, most of these primitives are either too large to fit on resource-constrained devices or highly inefficient when implemented on them. There have been many attempts to address this problem in the literature where cryptography engineers modify conventional primitives into lightweight versions or build new lightweight primitives from scratch. Unfortunately, both solutions suffer from either reduced security, low performance, or high implementation cost. This thesis investigates the performance of the conventional cryptographic primitives and explores the effect of their different building blocks and design choices on their performance. It also studies the impact of the various implementations approaches and optimisation techniques on their performance. Moreover, it investigates the limitations imposed by the tight processing and storage capabilities in constrained devices in implementing cryptography. Furthermore, it evaluates the performance of many newly designed lightweight cryptographic primitives and investigates the resources required to run them with acceptable performance. The thesis aims to provide an insight into the performance of the cryptographic primitives and the resource needed to run them with acceptable performance. This will help in providing solutions that balance performance, security, and resource requirements for these devices.