Evaluation of lightweight data integrity algorithms for the Internet of Things

Abstract

Advances in ubiquitous computing have led to the use of the Internet of Things (IoT). IoT interconnects various physical objects, such as computers, mobile phones, vehicles, and sensors, to the internet. In an IoT platform, sensors and actuators monitor, collect, and process the data from their environment. Since the IoT is made up of devices with limited resources, such as limited memory, computation power, and processing speed, primitive cryptographic algorithms are not practical for them. Due to this, lightweight cryptographic algorithms are used in IoT systems to provide the necessary security. The selection of a cryptographic algorithm is critical in a resource-constrained environment for improving performance and security. With the widespread use of resource-constrained devices that should be secure, it is difficult to development lightweight cryptographic hash functions that are able to meet security requirements while also providing good performance. A lightweight cryptographer/designer has to deal with the trade-off between security, cost, and performance. Only two of these, security and low costs, security and performance, or low costs and performance, can be easily implemented, but not all three at the same time. To balance the trade-off between security, cost, and performance, lightweight hash functions with smaller internal state sizes, smaller output sizes, and short message sizes should be used. With many lightweight algorithms being researched and developed, it is not easy for researchers and designers to know which one they should choose for their application. Performance evaluation of hash functions is necessary, as it updates the body of knowledge and gives academics a guideline for selecting lightweight hash functions that are secure and can perform well for their applications’ needs. This study aims to update the body of knowledge on the latency and memory consumption performance of two widely used lightweight hash functions called SPONGENT and PHOTON. Extensive research around SPONGENT, which is a family of thirteen variants, and PHOTON, which is a family of five variants, has been done. They are also listed as standardised lightweight hash functions in ISO/IEC 29192 Part 5. The Raspberry Pi 3 Model B was used to test and evaluate SPONGENT and PHOTON in two test cases: Test Case 1 analyses and evaluates the performance when the data log file used as input to the hash increases in size. Test case 2 studies how adapting the capacity/input rate affects performance.