Exploiting non-constant safe memory in resilient algorithms and data structures

We extend the Faulty RAM model by Italiano and Finocchi (2008) by adding a safe memory of arbitrary size S, and we then derive tradeoffs between the performance of resilient algorithmic techniques and the size of the safe memory. In particular, we describe a resilient algorithm for sorting n entries in O (n logn+ α(δ/S + logS)) time using Θ (S) safe memory locations, where α ≤ δ denotes the actual number of faults. Our algorithm outperforms previous resilient sorting algorithms which do not exploit the available safe memory and require O (n logn+ αδ) time. Finally, we exploit our merging algorithm for deriving a resilient priority queue data structure requiringO (logn+ δ/S) amortized time per operation. Our priority queue improves the O (logn+ δ) amortized time required by the state of the art.