An memristor-based synapse implementation using BCM learning rule

Yongchuang Huang; Junxiu Liu; Jim Harkin; Liam McDaid; Yuling Luo

doi:10.1016/j.neucom.2020.10.106

An memristor-based synapse implementation using BCM learning rule

Yongchuang Huang, Junxiu Liu, Jim Harkin, Liam McDaid, Yuling Luo

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

Abstract

A novel memristive synapse model based on the HP memristor is proposed in this paper, which can address the problem of synaptic weight infinite modulations. The sliding threshold mechanism of the Bienenstock-Cooper-Munro rule (BCM) is used to redefine the memristance (i.e. synaptic weight) adjustment process of the memristive synapse model. Based on the proposed memristor-based synapse and Leaky Integrate-and-Fire neurons, a spiking neural network (SNN) hardware fragment is constructed, where spike trains with different frequencies are used to evaluate the stability performance of the proposed SNN hardware. Results show that compared to other approaches, the network is stable under different stimuli due to the characteristics of the memristor-based synapse model, and prove that the proposed synapse model is able to mimic biological synaptic behaviour and the problem of synaptic weight infinite modulations is addressed.

Original language	English
Pages (from-to)	336-342
Number of pages	7
Journal	Neurocomputing
Volume	423
Early online date	16 Nov 2020
DOIs	https://doi.org/10.1016/j.neucom.2020.10.106
Publication status	Published - 29 Jan 2021

Bibliographical note

Funding Information:
This research is supported by the National Natural Science Foundation of China under Grant 61976063 and the funding of Overseas 100 Talents Program of Guangxi Higher Education.

Funding Information:
Liam J. McDaid received the B.Eng. (Hons.) degree in Electrical and Electronics Engineering and the Ph.D. degree in solid-state devices from the University of Liverpool, Liverpool, U.K., in 1985 and 1989, respectively. He is currently a Professor of Computational Neuroscience with Ulster University, U.K., and leads the Computational Neuroscience and Neural Engineering Research Team. His current research interests include modeling the role of glial cells in the functional and dysfunctional brain and he is also involved in the development of software/hardware models of neural-based computational systems, with particular emphasis on the mechanisms that underpin self-repair in the human brain. Dr. McDaid received several research grants in this domain and is currently a Collaborator on an HFSP and EPSRC funded project.

Publisher Copyright:
© 2020 Elsevier B.V.

Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.

Keywords

BCM theory
Learning rule
Memristor
Spiking neural networks

Access to Document

10.1016/j.neucom.2020.10.106

Cite this

@article{c3b582ee6f5b49c78ce6e7df9fa2ff7b,

title = "An memristor-based synapse implementation using BCM learning rule",

abstract = "A novel memristive synapse model based on the HP memristor is proposed in this paper, which can address the problem of synaptic weight infinite modulations. The sliding threshold mechanism of the Bienenstock-Cooper-Munro rule (BCM) is used to redefine the memristance (i.e. synaptic weight) adjustment process of the memristive synapse model. Based on the proposed memristor-based synapse and Leaky Integrate-and-Fire neurons, a spiking neural network (SNN) hardware fragment is constructed, where spike trains with different frequencies are used to evaluate the stability performance of the proposed SNN hardware. Results show that compared to other approaches, the network is stable under different stimuli due to the characteristics of the memristor-based synapse model, and prove that the proposed synapse model is able to mimic biological synaptic behaviour and the problem of synaptic weight infinite modulations is addressed.",

keywords = "BCM theory, Learning rule, Memristor, Spiking neural networks",

author = "Yongchuang Huang and Junxiu Liu and Jim Harkin and Liam McDaid and Yuling Luo",

note = "Funding Information: This research is supported by the National Natural Science Foundation of China under Grant 61976063 and the funding of Overseas 100 Talents Program of Guangxi Higher Education. Funding Information: Liam J. McDaid received the B.Eng. (Hons.) degree in Electrical and Electronics Engineering and the Ph.D. degree in solid-state devices from the University of Liverpool, Liverpool, U.K., in 1985 and 1989, respectively. He is currently a Professor of Computational Neuroscience with Ulster University, U.K., and leads the Computational Neuroscience and Neural Engineering Research Team. His current research interests include modeling the role of glial cells in the functional and dysfunctional brain and he is also involved in the development of software/hardware models of neural-based computational systems, with particular emphasis on the mechanisms that underpin self-repair in the human brain. Dr. McDaid received several research grants in this domain and is currently a Collaborator on an HFSP and EPSRC funded project. Publisher Copyright: {\textcopyright} 2020 Elsevier B.V. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.",

year = "2021",

month = jan,

day = "29",

doi = "10.1016/j.neucom.2020.10.106",

language = "English",

volume = "423",

pages = "336--342",

journal = "Neurocomputing",

issn = "0925-2312",

publisher = "Elsevier",

}

TY - JOUR

T1 - An memristor-based synapse implementation using BCM learning rule

AU - Huang, Yongchuang

AU - Liu, Junxiu

AU - Harkin, Jim

AU - McDaid, Liam

AU - Luo, Yuling

N1 - Funding Information: This research is supported by the National Natural Science Foundation of China under Grant 61976063 and the funding of Overseas 100 Talents Program of Guangxi Higher Education. Funding Information: Liam J. McDaid received the B.Eng. (Hons.) degree in Electrical and Electronics Engineering and the Ph.D. degree in solid-state devices from the University of Liverpool, Liverpool, U.K., in 1985 and 1989, respectively. He is currently a Professor of Computational Neuroscience with Ulster University, U.K., and leads the Computational Neuroscience and Neural Engineering Research Team. His current research interests include modeling the role of glial cells in the functional and dysfunctional brain and he is also involved in the development of software/hardware models of neural-based computational systems, with particular emphasis on the mechanisms that underpin self-repair in the human brain. Dr. McDaid received several research grants in this domain and is currently a Collaborator on an HFSP and EPSRC funded project. Publisher Copyright: © 2020 Elsevier B.V. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2021/1/29

Y1 - 2021/1/29

N2 - A novel memristive synapse model based on the HP memristor is proposed in this paper, which can address the problem of synaptic weight infinite modulations. The sliding threshold mechanism of the Bienenstock-Cooper-Munro rule (BCM) is used to redefine the memristance (i.e. synaptic weight) adjustment process of the memristive synapse model. Based on the proposed memristor-based synapse and Leaky Integrate-and-Fire neurons, a spiking neural network (SNN) hardware fragment is constructed, where spike trains with different frequencies are used to evaluate the stability performance of the proposed SNN hardware. Results show that compared to other approaches, the network is stable under different stimuli due to the characteristics of the memristor-based synapse model, and prove that the proposed synapse model is able to mimic biological synaptic behaviour and the problem of synaptic weight infinite modulations is addressed.

AB - A novel memristive synapse model based on the HP memristor is proposed in this paper, which can address the problem of synaptic weight infinite modulations. The sliding threshold mechanism of the Bienenstock-Cooper-Munro rule (BCM) is used to redefine the memristance (i.e. synaptic weight) adjustment process of the memristive synapse model. Based on the proposed memristor-based synapse and Leaky Integrate-and-Fire neurons, a spiking neural network (SNN) hardware fragment is constructed, where spike trains with different frequencies are used to evaluate the stability performance of the proposed SNN hardware. Results show that compared to other approaches, the network is stable under different stimuli due to the characteristics of the memristor-based synapse model, and prove that the proposed synapse model is able to mimic biological synaptic behaviour and the problem of synaptic weight infinite modulations is addressed.

KW - BCM theory

KW - Learning rule

KW - Memristor

KW - Spiking neural networks

UR - http://www.scopus.com/inward/record.url?scp=85097074920&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2020.10.106

DO - 10.1016/j.neucom.2020.10.106

M3 - Article

VL - 423

SP - 336

EP - 342

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -

An memristor-based synapse implementation using BCM learning rule

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Embargoed Document

Fingerprint

Cite this