ARTICLE IN PRESS JID: NEUCOM [m5G;May 21, 2018;18:26] Neurocomputing 000 (2018) 1–11 Contents lists available at ScienceDirect Neurocomputing journal homepage: www.elsevier.com/locate/neucom Implementation of adaptive neuron based on memristor and memcapacitor emulators Mohammad Saeed Feali, Arash Ahmadi ∗ , Mohsen Hayati Electrical Engineering Department, Razi University, Kermanshah, Iran a r t i c l e i n f o Article history: Received 10 June 2017 Revised 17 December 2017 Accepted 3 May 2018 Available online xxx Communicated by Duan Shukai Keywords: Neuromorphic Adaptive neuron Memristor Memcapacitor a b s t r a c t Adaptive response to timely constant stimuli is a common feature of biological neurons. Implementation of neurons with such features is important for achieving biologically plausible networks using electronic systems. Emerging memristor devices open new horizons in electronically implementation of neural net- works with high integration density and low power consumption. Promising potential applications can be considered for mem-elements (memristor, memcapacitor, and meminductor) with their built-in memory- properties. Since the dynamics of mem-elements makes them suitable for direct emulation of biological features of real neurons, it is expected that implementation of real neurons with complicated behavior to be straightforward using mem-element without complicating the implementation of the whole neuron circuit. Mem-elements are still unavailable commercially, so we utilize the mem-elements emulators to evaluate the feasibility of using these elements in the implementation of adaptive neurons. To ensure that there is the possibility of practical implementation of our neuron circuit, we used mem-elements emula- tors in SPICE simulations instead of mem-elements behavioral models. This work is among the first pa- pers in the implementation of adaptive neurons using mem-elements. Here, using memristor and mem- capacitor emulators the neuristor with adaptive behavior is implemented in SPICE environment. We use two different methods for induction of adaptive behavior to the neuristor response. In the first method, the capacitor in the primary circuit of neuristor is replaced with memcapacitor. Alternatively, the cou- pling resistor in the primary circuit of neuristor is replaced with the memristor in the second method. Results show that, the feature of memristor/memcapacitor in changing its resistance/capacitance dur- ing time upon excitation with current or voltage, makes the neuristor behavior to be adaptive in both methods, i.e. the neuristor shows the spike-frequency adaptation behavior in response to the continuous external stimulus, where the frequency of generated spikes depends on the duration of the external input stimulus. © 2018 Elsevier B.V. All rights reserved. 1. Introduction The human brain contains around 100 billion neurons and one quadrillion synapses [1]. Each neuron acts as a processing unit in the brain, which consists of soma, dendrites, and axon. Informa- tion from one neuron flows to another neuron across a synapse. Neurons communicate with each other by releasing action poten- tials in the form of spikes. Action potentials are generated by spe- cial types of voltage-gated ion channels embedded in a neuron’s membrane [2], causing the momentary change in electrical poten- tial on the surface of the neuron. Hodgkin and Huxley described a conductance-based model to explain the ionic mechanisms un- derlying the initiation and propagation of action potentials in the squid giant axon [3]. The spatial and temporal pattern of these ∗ Corresponding author. E-mail address: aahmadi70@gmail.com (A. Ahmadi). spikes represents information in neural networks [4,5]. The firing rate of neurons is related to the strength of the inputs stimuli in which stronger stimulation causes higher firing rate of the neuron. The recent history of neurons electrical activity affects the gen- erated spike train, known as spike-frequency adaptation [6]. This mechanism affects neuron activation in the brain, so that the fir- ing rate of neuron spikes is reduced during a sustained stimulus. It is believed that the existence of several different ion channels in the neuron with different time constants and thresholds are the main reason for spike-frequency adaptation. Several processes can produce the spike-frequency adaptation. In a number of silicon- based neurons, a mechanism is used to produce slow ionic cur- rents with each spike that are subtracted from the input current to the neuron [7]. This mechanism serves as a negative feedback, which is modeled differently in several silicon-based neurons [7]. In some silicon-based neurons, the effect of calcium-dependence, after-hyperpolarization potassium currents in real neurons, is https://doi.org/10.1016/j.neucom.2018.05.006 0925-2312/© 2018 Elsevier B.V. All rights reserved. Please cite this article as: M.S. Feali et al., Implementation of adaptive neuron based on memristor and memcapacitor emulators, Neu- rocomputing (2018), https://doi.org/10.1016/j.neucom.2018.05.006