A novel True Random Bit Generator design for image encryption

Abstract

Random Numbers are popular research topic due to the performance in encryption techniques. Owing to the protection of communication data against eavesdroppers, randomness is quiet important. The aim of this paper is to introduce an easily applicable True Random Bit Generator for image encryption. For this purpose, electromagnetic pollution is used as entropy source. Normalization procedure with very simple mathematical equations are applied on electric field strength values. Additionally, XOR post-processing procedure is performed. Different randomness test are studied for Random Numbers: NIST 800-22 Test Suite, Autocorrelation Test, Scale-Index Test. Subsequently, Tests are applied on normalized data and post-processed data separately. NIST Tests and Scale-Index Test are accomplished for both data, but Autocorrelation Test is successful for only Post-Processed values. Encryption of images are made with generated bits. This paper proves that, performing only NIST Tests are not trustworthy in True Random Bit Generators. High Frequency Electromagnetic Pollution firstly used as an entropy source in this work. Encryption process of sample images is concluded successfully. © 2019 Elsevier B.V.