Entropy analysis of data collected from inertial measurement unit of cyber-physical system under non-disturbed conditions

Abstract

Nowadays cyber-physical systems are widely used for many purposes. We consider the provision of information security of data channels in such systems. Cryptographic data security approach based on random sequences is commonly used to solve this task. Its reliability depends on quality of random data being used, thus truly random sequences are preferable for application. Truly random data generation is a time-consuming process and it requires entropy sources of physical nature. The goal of the paper presented is to research methods and approaches of collecting random numbers using inertial measurement unit as a part of cyber-physical system. Method. Quality assessment of a binary sequence was carried out during the research by determination of random sequence statistical characteristics.Main Results. Research results have shown up that raw data collected from onboard inertial sensors possess lack of entropy under non-disturbed conditions, therefore an additional post-processing is required. Practical Relevance. The results of the research can be used to obtain random sequences for on board cyber-physical systems equipped with inertial measurement units without the use of additional devices. It is planned to collect data from a flying unmanned aerial system in future to apply extractors and to utilize other methods in order to improve quality of a binary sequence.