Generating Random Numbers from Biological Signals in LabVIEW Environment and Statistical Analysis

Abstract

This paper explores the generation of random numbers, using electromyographic (EMG) signals collected from arm, elbow and finger movements of healthy individuals. The original signals were extracted from the Ninaweb database. The author designed a new discretization algorithm to convert these signals from floating point numbers to discrete values, and proposed a true random number generator (TRNG) structure that obtain the EMG signals with human arm and finger movements as noise sources. The proposed algorithm was applied to obtain and process real-time signals in the LabVIEW environment, and verified through NIST, TestU01, Scale index and autocorrelation tests. The results show that the discretization algorithms in TRNGs faced a huge data loss (70 %), while the designed algorithm with our structure lost no data and achieved 100 % efficiency in number generation. The research results prove the possibility of generating random numbers from biological signals.