Securing Image Privacy in Internet-of-Vehicles With a Multiwing Hyperchaotic Memristive Neural Network

In the Internet-of-Vehicles (IoV), transmission and storage of image data generated from traffic surveillance, vehicle monitoring, and driver assistance systems are becoming increasingly common. However, the sensitive and critical nature of such data present significant security and privacy challenges. To effectively address these issues, this paper introduces a novel and robust solution empowered by a multiwing hyperchaotic memristive neural network (MH-MNN). Drawing inspiration from the chaotic discharge patterns of the human brain, we develop a new neural network model incorporating highly non-linear memristive synapses. This model generates hyperchaotic sequences with complex multiwing topologies, offering a higher degree of randomness and unpredictability than traditional chaotic systems.The MH-MNN boasts an expanded parameter space and significantly enhances adjustability and complexity, making it a superior option for cryptographic applications. To validate its practical use, we implement the MH-MNN on a field-programmable gate array (FPGA) platform, applying it to image encryption in IoV communication. Extensive experimental results reveal that our method can dramatically improve resistance to various attacks while maintaining image integrity. The information entropy approaches the ideal value of 8, key sensitivity reaches 10⁻¹⁶, and the Number of Pixels Change Rule (NPCR) and the Unified Average Change Intensity (UACI) are approximately 99.60% and 33.44%, respectively. These results highlight the system’s exceptional suitability for resource-constrained IoV environments, where both high security and efficiency are paramount.