Effect of Device Mismatches in Differential Oscillatory Neural Networks

Analog implementation of Oscillatory Neural Networks (ONNs) has the potential to implement fast and ultra-low-power computing capabilities. One drawbacks analog is component mismatches which cause desynchronization instability in ONNs. Emerging devices like memristors VO2are particularly prone variations. In this paper, we study effect on performance differential ONNs (DONNs). Mismatches were considered two main blocks: oscillatory neurons synaptic circuits. To measure DONN tolerance each block, was evaluated with being present separately block. Memristor-bridge circuits four used as The based VO2-devices. simulation results showed that more vulnerable components than DONNs found tolerate up 20% memristance However, resulted non-uniformity natural frequencies, causing instability. Simulations 0.5% relative standard deviation (RSD) frequencies can reduce dramatically. addition, sensitivity analyses high threshold voltage VO2-devices most sensitive parameter for frequency desynchronization.