Experimental validation of IoT image encryption scheme based on a 5-D fractional hyperchaotic system and Numba JIT compiler

In the realm of Internet of Things (IoT) systems, the interconnectivity of physical hardware devices is a fundamental aspect, and as a result, data exchange assumes a critical role in the network. Given the sensitivity of such information, it is imperative to adopt appropriate measures to encrypt the data to safeguard it from unauthorized access. It is, therefore, paramount to prove novel encryption algorithms at an experimental level. To overcome that, a potential solution is fractional-order multi-scroll chaotic systems, which provide an extra degree of freedom to enhance ergodicity and random-like behaviors, which can help to improve encryption keyspace to get robust ciphers. However, the numerical algorithms to obtain a fractional chaotic series increase the computational cost because they demand extensive simulation time to model the memory of the standard fractional derivatives, limiting highspeed encryption. In this framework, a fast encryption scheme using a 5D fractional-order (FO) hyper-chaotic multi-scroll (HCMS) system is proposed and verified experimentally. Based on multiprocessing strategies and the Numba just -in -time (JIT) compiler, the Python code that describes the FO-HCMS system is optimized, which enables image encryption in real -time. The physical implementation of the encryption scheme is performed on an Advance RISC Machine (ARM) processor, and it is applied for image encryption on a machine-to-machine (M2M) communication using the message queuing telemetry transport (MQTT) protocol. The obtained results indicate that the proposed encryption scheme can reach encryption throughputs of up to 19.891 Mbps on an ARM with a 1.4 GHz CPU and 77.864 Mbps on a PC with a 3.1 GHz CPU.