Abstract
Neural networks are achieving state-of-the-art performance in many applications, from speech recognition to computer vision. A neuron in a multi-layer network needs to multiply each input by its weight, sum the results and perform an activation function. This paper is an extended version of the article in which we present an implementation of an amplifier-based MOS analog neuron and the optimization of the synaptic weights using in-loop circuit simulations. In addition to the base topology, we present two variations of the original conference paper topology to reduce area and power. MOS transistors operating in the triode region are used as variable resistors to convert the input and weight voltage to proportional input current. To test the analog neuron in full networks, an automatic generator is developed to produce a netlist based on the number of neurons on each layer, inputs, and weights. Simulation results using a CMOS 180 nm technology for all topologies demonstrate the neuron proper transfer function and its functionality while trained in test datasets.
Similar content being viewed by others
References
Janardan Misra, I. S. (2010). Artificial neural networks in hardware: A survey of two decades of progress. Neurocomputing, 74, 239–255.
Baibhab Chatterjee Priyadarshini Panda, S. M. A. B. K. R., & Sen, S. (2018). Exploiting inherent error-resiliency of neuromorphic computing to achieve extreme energy-efficiency through mixed-signal neurons. arXiv preprint.
Buhler, F. N., Brown, P., Li, J., Chen, T., Zhang, Z., & Flynn, M. P. (2017). A 3.43tops/w 48.9pj/pixel 50.1nj/classification 512 analog neuron sparse coding neural network with on-chip learning and classification in 40 nm CMOS. In 2017 Symposium on VLSI circuits (pp. C30–C31).
Salam, F. M. A., Choi, M. R., & Wang, Y. (1990). An analog MOS implementation of the synaptic weights for feedforward/feedback neural nets. In Proceedings of the 32nd Midwest symposium on circuits and systems.
Wang, Y., & Salam, F. M. A. (1990). Design of neural network systems from custom analog VLSI chips. In IEEE international symposium on circuits and systems.
Choi, M. R., & Salam, F. M. (1991). Implementation of feedforward artificial neural nets with learning using standard CMOS VLSI technology. In IEEE international symposium on circuits and systems.
Salam, F. M. A., & Choi, M. R. (1991). Analog MOS vector multipliers for the implementation of synapses in artificial neural networks. Journal of Circuits, Systems and Computers, 1, 205–228.
Salam, F., Wang, Y., & Oh, H. J. (1999). A 50-neuron CMOS analog chip with on-chip digital learning: Design, development, and experiments. Computers & Electrical Engineering, 25(5), 357–378.
EI-Bakry, M. H., & Mastorakis, N. (2009). A simple design and implementation of reconfigurable neural networks. In Proceedings of international joint conference on neural networks.
Larras, B., Lahuec, C., Seguin, F., & Arzel, M. (2016). Ultra-low-energy mixed-signal ic implementing encoded neural networks. IEEE Transactions on Circuits and Systems I: Regular Papers, 63(11), 1974–1985.
Gatet, L., Tap-Beteille, H., & Lescure, M. (2008). Analog neural network implementation for a real-time surface classification application. IEEE Sensors Journal, 8(8), 1413–1421.
Ngwar, M., & Wight, J. (2015). A fully integrated analog neuron for dynamic multi-layer perceptron networks. In 2015 International joint conference on neural networks (IJCNN).
Szczesny, S. (2017). 0.3 v 2.5 nw per channel current-mode CMOS perceptron for biomedical signal processing in amperometry. IEEE Sensors Journal, 17, 5399–5409.
Weber, T. O., da Silva Labres, D., & Cabrera, F. L. (2019). Amplifier-based MOS analog neural network implementation and weights optimization. In Proceedings of the 32nd symposium on integrated circuits and systems design (SBCCI’19). New York, NY, USA: Association for Computing Machinery.
Kirkpatrick, S. C. D., Gelatt, J., & Vecchi, M. P. (1983). Optimization by simulated annealing. Science, 220, 671–680.
Fisher, R. A. (1936). The use of multiple measurements in taxonomic problems. Annals of Eugenics, 7(2), 179–188.
Khodabandehloo, G., Mirhassani, M., & Ahmadi, M. (2012). A prototype CVNS distributed neural network using synapse-neuron modules. IEEE Transactions on Circuits and Systems I: Regular Papers, 59(7), 1482–1490.
Youssefi, B., Leigh, A. J., Mirhassani, M., & Wu, J. (2018). Hardware realization of mixed-signal neural networks with modular synapson–neuron arrays. In 2018 IEEE international symposium on circuits and systems (ISCAS).
Geng, C., Sun, Q., & Nakatake, S. (2020). An analog CMOS implementation for multi-layer perceptron with relu activation. In 2020 9th International conference on modern circuits and systems technologies (MOCAST) (pp. 1–6).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Oliveira Weber, T., Cabrera, F.L. & da Silva Labres, D. Topology Variations of an Amplifier-based MOS Analog Neural Network Implementation and Weights Optimization. Analog Integr Circ Sig Process 106, 635–647 (2021). https://doi.org/10.1007/s10470-021-01798-y
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10470-021-01798-y