中国物理B ›› 2020, Vol. 29 ›› Issue (2): 28502-028502.doi: 10.1088/1674-1056/ab65b5

• INTERDISCIPLINARY PHYSICS AND RELATED AREAS OF SCIENCE AND TECHNOLOGY • 上一篇    下一篇

Memristor-based vector neural network architecture

Hai-Jun Liu(刘海军), Chang-Lin Chen(陈长林), Xi Zhu(朱熙), Sheng-Yang Sun(孙盛阳), Qing-Jiang Li(李清江), Zhi-Wei Li(李智炜)   

  1. College of Electronic Science, National University of Defense Technology, Changsha 410073, China
  • 收稿日期:2019-09-25 修回日期:2019-11-11 出版日期:2020-02-05 发布日期:2020-02-05
  • 通讯作者: Zhi-Wei Li E-mail:lizhiwei@nudt.edu.cn
  • 基金资助:
    Project supported by the National Natural Science Foundation of China (Grant Nos. 61471377, 61804181, 61604177, and 61704191).

Memristor-based vector neural network architecture

Hai-Jun Liu(刘海军), Chang-Lin Chen(陈长林), Xi Zhu(朱熙), Sheng-Yang Sun(孙盛阳), Qing-Jiang Li(李清江), Zhi-Wei Li(李智炜)   

  1. College of Electronic Science, National University of Defense Technology, Changsha 410073, China
  • Received:2019-09-25 Revised:2019-11-11 Online:2020-02-05 Published:2020-02-05
  • Contact: Zhi-Wei Li E-mail:lizhiwei@nudt.edu.cn
  • Supported by:
    Project supported by the National Natural Science Foundation of China (Grant Nos. 61471377, 61804181, 61604177, and 61704191).

RichHTML

0

PDF (PC)

143

赞

摘要： Vector neural network (VNN) is one of the most important methods to process interval data. However, the VNN, which contains a great number of multiply-accumulate (MAC) operations, often adopts pure numerical calculation method, and thus is difficult to be miniaturized for the embedded applications. In this paper, we propose a memristor based vector-type backpropagation (MVTBP) architecture which utilizes memristive arrays to accelerate the MAC operations of interval data. Owing to the unique brain-like synaptic characteristics of memristive devices, e.g., small size, low power consumption, and high integration density, the proposed architecture can be implemented with low area and power consumption cost and easily applied to embedded systems. The simulation results indicate that the proposed architecture has better identification performance and noise tolerance. When the device precision is 6 bits and the error deviation level (EDL) is 20%, the proposed architecture can achieve an identification rate, which is about 92% higher than that for interval-value testing sample and 81% higher than that for scalar-value testing sample.

关键词: memristor, memristive devices, vector neural network, interval

Abstract: Vector neural network (VNN) is one of the most important methods to process interval data. However, the VNN, which contains a great number of multiply-accumulate (MAC) operations, often adopts pure numerical calculation method, and thus is difficult to be miniaturized for the embedded applications. In this paper, we propose a memristor based vector-type backpropagation (MVTBP) architecture which utilizes memristive arrays to accelerate the MAC operations of interval data. Owing to the unique brain-like synaptic characteristics of memristive devices, e.g., small size, low power consumption, and high integration density, the proposed architecture can be implemented with low area and power consumption cost and easily applied to embedded systems. The simulation results indicate that the proposed architecture has better identification performance and noise tolerance. When the device precision is 6 bits and the error deviation level (EDL) is 20%, the proposed architecture can achieve an identification rate, which is about 92% higher than that for interval-value testing sample and 81% higher than that for scalar-value testing sample.

Key words: memristor, memristive devices, vector neural network, interval

中图分类号:  (Nanoelectronic devices)

  • 85.35.-p
87.19.lv (Learning and memory) 07.05.Mh (Neural networks, fuzzy logic, artificial intelligence)