Biosignal encryption algorithm based on Ushio chaotic map for e-health

In last years, chaos-based encryption algorithms have been proposed in the literature to provide information confidentiality in digital images, biometric systems, telemedicine, multiuser networks, among others. In e-health, the confidentiality of biomedical signals is required to avoid illegal activities such as identity theft, wrong medical prescription, fraud or extortion against patients or the healthcare entities. On the other hand, improve the randomness of the chaotic map can increase the randomness of the cryptograms and thus, increase the security of the whole cryptosystem. In this paper, we first propose to improve the randomness of five selected chaotic maps by using trigonometric (sine, cosine and tangent) and the exponential function combined with module one operation. Then, we use the monobit test of NIST 800-22 suite to determine the level of achieved randomness. Finally, we use the selected chaotic map in an encryption algorithm for clinical signals based on improved sequences of two Ushio maps and just one round of confusion-diffusion process. In experimental results, we use electrocardiogram (ECG), electroencephalogram (EEG), and blood pressure signals from PhysioBank ATM data base at different sampling frequency and records of 10 seconds. Based on the simulation results in MATLAB, the main findings are (1) the 2D Ushio map with exponential function and module one achieves the best randomness with 0.9760 in P_value; (2) the encryption scheme presents high robustness against known attacks such extreme sensitivity to plain biosignal and the 128-bit secret key, correlation between plain biosignal and cryptogram close to 0, autocorrelation of cryptogram close to 0, floating frequency of 82.4%, uniform histograms in cryptograms, robustness against noise and occlusion attacks, and high speed of encryption; (3) the cryptosystem can be implemented in (low-cost) limited processing systems such as microcontrollers for wireless body area networks (WBAN) applications in e-health; (4) the propose encryption approach can be used to transmit medical signals securely in e-health for monitoring or diagnosis of diseases.