Securing Personal and Body Area Networks Based on Physiological Datasets

Securing Personal and Body Area Networks Based on Physiological Datasets

Jagdish Panwar1, Prof. Pankaj Raghuwanshi2

Abstract: As BANs have limited processing power and memory availability, securing body area networks or body sensor networks is problematic. Due to the impracticality of implementing high-end encryption in resource-constrained body sensor networks, heartbeat-based security is preferable. To simulate real heartbeats from electrocardiogram (ECG) data, the suggested method uses a deep Markov model to generate random bit sequences (RBS). The interpulse interval (IPI) is defined by the information retrieved from the RR interval, the SS interval, and the QRS complex. The MIT-BIH database (MIT-BIHdb) is the source of the extracted data. Both the entropy and the hamming distance are used to evaluate the performance of the suggested method. Results demonstrate that, in comparison to prior work, the suggested method achieves a greater hamming distance for the amount of bits extracted per IPI. The entropy is somewhat greater than the prior method, fluctuating between 0.95 and 1. The receiving end's ability to accurately and distinguishably detect the received binary sequence for authorization is shown by the rise in hamming distance..

Keywords: Body Sensor Networks, Inter Pulse Interval (IPI), Random Bit Stream (RBS), Markov Discrete Time Markov Chain, Hamming Distance, Entropy.