Resource-Efficient One-Dimensional Discrete Chaotic Map-Based Pseudo-Random Number Generator for IoT Applications: A Practical Analysis

Resource-Efficient One-Dimensional Discrete Chaotic Map-Based Pseudo-Random Number Generator for IoT Applications: A Practical Analysis

연구 분야: Verification

학회: SN Computer Science

Smart Internet of Things (IoT) applications are transforming lives by enabling devices to communicate and automate tasks seamlessly. These applications improve efficiency and user experience, but they also bring new security challenges, especially in environments with limited resources. In this context, cryptography plays a crucial role with a vital component being secure and resource-efficient Pseudo-Random Number Generators (PRNGs), which are essential for generating unpredictable encryption keys to safeguard communications and protect sensitive data. Conventional PRNGs are often unsuitable for IoT environments due to their complexity and resource-intensive nature. Chaotic maps, particularly one-dimensional (1D) maps, offer a lightweight alternative to conventional PRNGs for resource-constrained IoT devices. This study addresses the challenge of designing a resource-efficient PRNG tailored for IoT devices by leveraging one-dimensional chaotic maps. The paper introduces a lightweight, high-speed, low-power, and memory-efficient PRNG and presents a comprehensive comparison of its performance utilizing 21 chaotic maps including Chebyshev, Fibonacci Chaotification Model-Based Chebyshev, Coupled Sine, Cubic, Cubic Logistic, JSMP, Logistic, Fibonacci Chaotification Model-Based Logistic, Logistic May, Quadratic, Fibonacci Chaotification Model-Based Quadratic, Renyi map, Sine-Sinh-Sine, Sine, Fibonacci Chaotification Model-Based Sine, Singer, Fibonacci Chaotification Model-Based Singer, Simple Quadratic, Fibonacci Chaotification Model-Based Simple Quadratic, Squared Sine Logistic, and Tent maps. The proposed approach is distinguished by its lightweight design, providing superior efficiency and adaptability for low-power, memory-constrained applications. The PRNG is assessed in MATLAB for operation count, bit generation time, and performance, and evaluated against the 15 NIST statistical tests. Results demonstrate the PRNG to have low operation count, high speed, and significantly higher throughput compared to existing PRNGs, while successfully passing all NIST tests, demonstrating its efficiency and robustness. Notably, the PRNG is evaluated on the state-of-the-art ARM Cortex-M3-based LPC 1768 IoT development board. The key performance assessment metrics include current, voltage, power, energy, execution time, and RAM and ROM memory requirements. The results demonstrate that the proposed PRNG is exceptionally well-suited for IoT applications. Specifically, Quadratic and FCM Sine map-based PRNGs are optimally aligned with the requirements of low-power and battery-driven IoT devices. Conversely, Quadratic and Simple Quadratic map-based PRNGs provide better solutions for scenarios where minimal memory usage is paramount. Also, for time-critical IoT applications, Quadratic, and Tent map-based PRNGs provide faster key generation, ensuring efficient operation in critical scenarios. Thus, this paper underscores the potential of the proposed 1D discrete chaotic map-based PRNG in enhancing IoT security, offering an optimal balance between performance and resource utilization.