Hardware Design for Secure Telemedicine Using A Novel Framework, A New 4D Memristive Chaotic Oscillator, and Dispatched Gray Code Scrambler

Abstract

This study introduces an energy-efficient FPGA-based image encryption mechanism utilizing a 4D memristive chaotic oscillator and a Dispatched Gray Code Scrambler (DGCS) within a MATLAB/Simulink FPGA-in-the-loop framework. Tailored for secure telemedicine, the system improves confusion and diffusion via structured pixel scrambling and chaos-driven key generation. Security assessments indicate substantial robustness, with global entropy of 7.9973, local entropy of 7.9040, near-zero correlation coefficients, NPCR of 99.6170%, and UACI of 33.3172%. The system records a PSNR of 29.72 dB under 1% salt-and-pepper noise, and 19.76 dB under Gaussian noise with variance 0.001, showcasing considerable resilience to both impulsive and distributed distortions. This robustness against Gaussian noise is particularly vital in telemedicine, where image integrity is essential amidst transmission challenges. The keystream successfully passes NIST SP 800-22 and TestU01 statistical evaluations. Designed on an Artix-7 FPGA, the system's power consumption stands at 105 mW, utilizing 11.38% of LUTs, 6.25% of DSPs, and 10.48% of I/Os, achieving a performance frequency of 7.24 MHz. These findings underscore its appropriateness for embedded, low-latency, and noise-resistant image safeguarding in resource-limited medical settings.