Memristor-Based Neural Network Circuit of Long-term Memory

As an important part of memory, memristor-based neural networks of memory are seldom involved in long-term memory. In this paper, a memristor-based neural network of long-term memory is designed. The designed circuit consists of a synapse module and a voltage adjustment module. The functions such as learning, forgetting, long-term memory and self-protection are implemented by the circuit. When a neuron forms a long-term memory for a stimulus, it responds more quickly to the same stimulus again. The research of long-term memory is helpful for neural network to simulate more real brain memory. The neural network circuit of long-term memory provides a reference for the further development of brain-like systems.