Understanding switching variability and random telegraph noise in resistive RAM

A deeper understanding of the noise and variability sources in resistive switching memory (RRAM) is needed for device improvement and scaling down. To meet this challenge, this work addresses switching statistics and read noise in RRAM. First, an analytical model for resistance variability in set and reset states is presented. Then, random telegraph noise (RTN) is discussed in terms of trap-induced depletion in the conductive filament (CF) The model accounts for size-dependent RTN, while bias-dependent RTN is explained by Poole-Frenkel (PF) transport and Joule heating in atomic-size CFs.