UDC 621 SIMULINK MODEL OF SPIKING NEURAL OSCILLATOR Sumanjit Sen Gupta Department of Electronics & Communication Engineering, Academy of Technology, Hooghly, India Subhra Kanti Das Department of Robotics & Automation CSIR-CMERI, Durgapur, India Abstract. <...> A Simulink model of a spiking neural network of two neurons that works as a neural oscillator is presented. <...> Each neuron is based upon Izhikevich Spiking Neuron Model. <...> Average Response Model is used to model various aspects of synaptic transmissions. <...> The parameters in the mathematical differential equation describing each neuron are set such that they exhibit a regular spiking pattern of cortical neurons. <...> The duty cycle and frequency of oscillations of our oscillator are quite flexible and are tuned by varying one or more of a few parameters, for example, changing the nature of synapses. <...> The presence of very few parameters reduces the complexity of simulation and allows this neural oscillator model to be easily implemented in myriad applications that involve sustained rhythmic patterns. <...> An initial spike of stimulus is enough to drive this oscillator. <...> The number of continuous action potentials fired depends upon the intensity and duration of the stimulus, as well as on the inherent traits of the neuron [1, 2]. <...> The neuron cells communicate among each other via these action potentials that travel from one neuron to another through synaptic clefts. <...> A group of neurons connected by synapses work together to form a neural network. <...> A neural network performs innumerable functions and one of those is oscillatory behavior that assist various rhythmic movements like locomotion, respiration, feeding and mating patterns. <...> Most of these rhythmic patterns are generated without the interception of any particular extrinsic stimulus [3]. <...> Recent studies have revealed the involvement of neural oscillations in memory formation, adaptation to environment besides sensory activities like visual perception, olfaction, etc. <...> Motor activities like animal locomotion are controlled by a network of neural oscillators, called Central Pattern Generator (CPG) [4], situated in ganglion or spinal cord. <...> The silicon neural system represents an artificial neural system that uses realistic principles of neural computation and are more biophysically accurate than existing classical electronic neural network models owing to the behavioral similarity of semiconductor response (I-V Curves) with that of neurons. <...> The first <...>