Modeling and analysis of a phenomenological memtransistor and its digital realization within a 4D hyperchaotic Lorenz system

Abstract The von Neumann bottleneck is putting increasing pressure on standard computing architectures. This situation is encouraging research into new devices, including memtransistors, which are used in in-memory and neuromorphic computing. However, building physical devices is difficult, so practical mathematical models are needed for system-level design and analysis. This study addresses this need by proposing a phenomenological memtransistor model based on the well-known MOSFET framework. The key features of the model, including its hysteretic and neuromorphic behaviors, are first explained and carefully confirmed through numerical simulations. After this validation, the model is incorporated into the classical Lorenz system to create a new four-dimensional (4D) memtransistor-based dynamical system. Detailed investigations show that this system exhibits strong hyperchaotic behavior, supported by the presence of two positive Lyapunov exponents. The final stage of this research is the digital implementation and experimental verification of the complete 4D hyperchaotic system on an STM32 microcontroller platform. The close agreement between the experimental results and the numerical simulations confirms the effectiveness of the overall approach and shows that it provides a practical way to design and implement advanced memristive systems.