Key points are not available for this paper at this time.
Resistive-switching memory (RRAM) is attracting a considerable interest for the development of high-density nonvolatile memories. However, several scaling and reliability issues still affect the development path of RRAM. This work addresses random telegraph-signal noise (RTN) of the RRAM current, potentially affecting the memory stability. We show a clear resistance dependence of the RTN amplitude, and we propose a physical model describing the interaction of the localized current with a fluctuating defect. By estimating the diameter of the conductive filament, the model quantitatively accounts for the observed RTN amplitude, thus allowing for an analytical prediction of state stability in RRAMs.
Ielmini et al. (Mon,) studied this question.