Edinburgh Napier University

  This paper delves into a comprehensive analysis of fault-tolerant memory systems, focusing on recovery techniques modelled using Markov chains to address transient errors. The study revolves around the application of scrubbing methods in conjunction with Single Error Correction and Double Error Detection (SEC-DED) codes. It explores three primary models: 1) Exponentially distributed scrubbing, involving periodic checks of memory words within exponentially distributed time intervals; 2) Deterministic scrubbing, featuring regular, periodic word checks; and 3) Mixed scrubbing, which combines both probabilistic and deterministic scrubbing approaches. The research encompasses the estimation of reliability and Mean Time to Failure (MTTF) values for each model. Notably, the findings highlight the superior performance of mixed scrubbing over simpler scrubbing methods in terms of reliability and MTTF.