Solving XOR in Spike Neural Network (SNN) with Component-off-the-Shelf

This paper addresses the solution of the XOR problem with Spiking Neural Networks (SNN) in order to improve energy efficiency and computational performance as Moore's Law approaches its limits. SNN is capable of solving nonlinear problems while saving energy by mimicking the working principles of biological neurons. For this purpose, a SNN consisting of 12 neurons was implemented on a breadboard using the Leaky Integrate and Fire (LIF) model. In the input layer of the network, 50 Hz and 100 Hz signals are processed with frequency sensitive filters. With the help of bandpass and low-pass filters, additive and inverting operational amplifiers, the XOR problem is successfully solved.