IEEE

Transient Modeling of Nonlinear Components using Global Gated Deep Recurrent Neural Network Masked With Dropout

Abstract: This article introduces a novel approach for nonlinear electronic circuit modeling called Global Gated Deep Recurrent Neural Network (GGDRNN). GGDRNNs leverage the stacking of multiple ...
IEEE

Channel Prediction Using Deep Recurrent Neural Network With EVT-Based Adaptive Quantile Loss Function

Abstract: Ultra-reliable low latency communication (URLLC) systems are pivotal for applications demanding high reliability and low latency, such as autonomous vehicles. In such contexts, channel ...