DISEÑO COMPORTAMENTAL DE NEURONAS DE IMPULSO EN HARDWARE RECONFIGURABLE BASADAS EN EL MODELO DE IZHIKEVICH (BEHAVIORAL DESIGN OF SPIKING NEURONS IN RECONFIGURABLE HARDWARE BASED ON IZHIKEVICH MODEL)
Resumen
Los modelos matemáticos de neuronas han logrado ser computacionalmente eficientes y biológicamente plausibles, como es el modelo de Izhikevich. La implementación de neuronas artificiales en hardware se ha extendido en áreas como: control, robótica, entre otros. Sin embargo, migrar un modelo a hardware en un sistema reconfigurable se complica debido a la alta ocupación de recursos lógicos. En este trabajo se presenta el diseño de dos neuronas de impulsos que replican el comportamiento del modelo de Izhikevich y la propuesta de modificación de estas para una funcionalidad más amplia. Se realizan simulando el modelo como sistema dinámico en software con un tiempo de 100 ms, para obtener los valores de iteraciones en los que se generan impulsos. Son implementadas en lenguaje de descripción de hardware utilizando máquinas de estado. Como resultados se obtienen neuronas que ocupan 10 y 13 Flip-Flops de 1536 disponibles en una FPGA Spartan 3.
Palabras Clave: FPGA’s, Modelo de Izhikevich, Neurona de impulso.
Abstract
Mathematical models of neurons have managed to be computationally efficient and biologically plausible like Izhikevich's model. The implementation of artificial neurons in hardware has spread in areas such as: control, robotics, among others. However, migrating a model to hardware in a reconfigurable system is complicated due to the high occupation of logical resources. This work presents the design of two spiking neurons that replicate the behavior of the Izhikevich model and the proposal to modify these for a broader functionality. They are carried out by simulating the model as a dynamic system in software with a time of 100 ms, to obtain the values of iterations in which spikes are generated. They are implemented in hardware description language using state machines. The results are neurons that occupy 10 and 13 Flip Flops out of 1536 available in a Spartan 3 FPGA.
Keywords: FPGA’s, Izhikevich model, Spiking neurons.
Texto completo:
645-661 PDFReferencias
Diehl P., Zarrela G., Cassidy A. (2016), Conversion of Artificial Recurrent Neural Networks to Spiking Neural Networks for Low-power Neuromorphic Hardware arXiv:1601.04187v1 [cs.NE], 56-58.
Han J., Li, Z., Zheng, W., & Zhang, Y. (2020), Hardware implementation of spiking neural networks on FPGA. Tsinghua Science and Technology, 25(4), 479–486. https://doi.org/10.26599/tst.2019.9010019.
Izhikevich E. (2003). Simple model of spiking neurons. IEEE Transactions
on Neural Networks, 14(6), 1569–1572. https://doi.org/10.1109/tnn.2003.820440.
Izhikevich E. (2004). Which Model to Use for Cortical Spiking Neurons? IEEE
Transactions on Neural Networks, 15(5), 1063–1070.
https://doi.org/10.1109/tnn.2004.832719.
Izhikevich E. (2021). Eugene M. Izhikevich old (NSI) home page. Retrieved 27 May 2021, from http://www.izhikevich.org.
Kelemen J. (2007), From Artificial Neural Networks to Emotion Machines with Marvin Minsky. Acta Polytechnica Hungarica, Vol. 4 (No. 4), 5–7.
Kousanakis E., Dollas A. Sotiriades E. (2017), An Architecture for the Acceleration of a Hybrid Leaky Integrate and Fire SNN on the Convey HC-2ex FPGA-Based Processor. IEEE 25th Annual International Symposium on Field-Programmable Custom Computing Machines, 56-63.
Lammie C., Hamilton T., Azghadi M., (2018), Unsupervised Character Recognition with a Simplified FPGA Neuromorphic System. IEEE International Symposium on Circuits and Systems. 1-4. doi. 10.1109/ISCAS.2018.8351532.
Pearson J., Pipe G., Mitchinson B., Gurney K., Melhuish C., Gilhespy I., Nibouche M. (2007), Implementing Spiking Neural Networks for Real-Time Signal-Processing and Control Applications: A Model-Validated FPGA Approach. IEEE Transactions on Neural Networks, 18(5), 1472–1487. https://doi.org/10.1109/tnn.2007.891203.
Rice K., Bhuiyan M., Taha T. (2009), FPGA Implementation of Izhikevich Spiking Neural Networks for Character Recognition. International Conference on Reconfigurable Computing and FPGAs. 451-455.
Rocke P., McGinley B., Morgan F., Maher J. (2007), Reconfigurable Hardware Evolution Platform for a Spiking Neural Network Robotics Controller, Springer-Verlag Berlin Heidelberg, 373-737.
Salamanca A. (2017) Diseño e implementación de una neurona de impulsos en hardware reconfigurable [Tesis de maestría, Universidad de Guadalajara]. Repositorio institucional- Universidad de Guadalajara.
Sandoval-Ruiz C. (2017), VHDL Model of configurable neural networks applied to decoding in cognitive radio, Revista Ingeniería UC, Vol. 24, No. 3, 290-301.
URL de la licencia: https://creativecommons.org/licenses/by/3.0/deed.es
Pistas Educativas está bajo la Licencia Creative Commons Atribución 3.0 No portada. 
TECNOLÓGICO NACIONAL DE MÉXICO / INSTITUTO TECNOLÓGICO DE CELAYA
Antonio García Cubas Pte #600 esq. Av. Tecnológico, Celaya, Gto. México
Tel. 461 61 17575 Ext 5450 y 5146
pistaseducativas@itcelaya.edu.mx