A comparison of binary number generators for wireless body sensor networks security

Abstract

Binary numbers (BNs) generation is crucial for wireless body sensor networks (WBSNs) security because they can be utilized for authentication or encryption of physiological data in WBSNs for healthcare applications. The paper compared pseudorandom number generators and biometrics based binary number generators for WBSNs, i.e. generated from the linear congruential generator (LCG) and the interpulse intervals (IPIs) based generator (IBG), e.g. IPIs extracted from electrocardiogram (ECG) and photoplethysmogram (PPG). The results of the study found that the randomness of binary numbers derived from the two generators hardly have difference and the generation time and memory occupation of a 128-bit BN derived from the LCG and IBG is 435.14ms and 22.62ms, 1.236KB and 0.199KB respectively. The study opens up a few key issues for future investigation, including the initial seed generation for the LCG, IPIs extraction for the IBG and IPIs encoding schemes.