Implementing bionic associate memory based on spiking signal
- Publisher:
- ELSEVIER SCIENCE INC
- Publication Type:
- Journal Article
- Citation:
- Information Sciences, 2023, 649
- Issue Date:
- 2023-11-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 07Implementing bionic associate memory based on spiking signal.pdf | Published version | 1.91 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Most of the associate memory circuits are at the simulation stage. If these designs are to be realized in hardware, they pose substantial requirements in terms of experimental conditions. To address this challenge, the utilization of spiking signals for information transmission emerges as a potential solution. In this paper, a multifunctional associate memory circuit based on spiking signal is proposed and implemented in hardware. The circuit mainly consists of neuron circuits and synapse modules. These neuron circuits emulate the transmission of spiking signals akin to biological neurons. Fundamental associative memory functions, such as learning and forgetting, are achieved by modulating synaptic weights. This modulation is governed by the voltage differential and input signal sequencing within the circuit. Notably, leveraging parameters derived from simulations, the implementation of the associate memory circuit in hardware obviates the need for stringent experimental prerequisites. And the experimental results are consistent with the simulation results. Furthermore, the circuit is endowed with the capability to execute generalization and differentiation functions, enabling the discrimination between signals labeled as “bell 1200 Hz” and “bell 1500 Hz”. This work may advance the development of bio-inspired circuit based on memristor.
Please use this identifier to cite or link to this item: