Invariants of noise in cyber-physical systems components

Abstract

The article is devoted to the invariant of internal electrical noise of electronic devices, which are components of cyber-physical systems. Time series of noise signals show chaotic behavior. Invariants are based on the autocorrelation function of dynamic time series. Insignificant differences on the micro-level devices lead to changes in the dynamics of time series. It is shown that the form of the autocorrelation function is unchanged for each electronic device of the cyber-physical system. The dynamic authentication algorithm has been developed, which consists of choosing a range of time series, defining and calculating invariants, making decisions about authentication. The result of the operation of the algorithm can be transferred to the executive mechanism, depending on the practical problems in cyber-physical systems. Also for the pseudorandom sequence of the embedded program generator, the following values are predicted on the basis of invariants. Estimated errors are calculated.