Deep Learning Based Detection for Mitigating Sneak Path Interference in Resistive Memory Arrays

In the resistive random access memory (RRAM) crossbar array, the sneak path problem severely degrades the reliability of data stored in the memory cell. In this paper, we apply the deep learning technique and propose a deep neural network (DNN) architecture together with a novel data preprocessing scheme for signal detection of RRAM in the presence of the sneak path interference. Simulation results demonstrate that our proposed DNN based detector with data preprocessing can effectively mitigate the sneak path effect without prior knowledge of the channel.