A range of attacks on network components, such as algorithmic denial-of-service attacks and cryptanalysis via timing attacks, are enabled by data structures for which an adversary can predict the durations of operations that he will induce on the data structure. In this paper we introduce the problem of designing data structures that confound an adversary attempting to predict the timing of future operations he induces, even if he has adaptive and exclusive access to the data structure and the timings of past operations. We also design a data structure for implementing a set (supporting membership query, insertion, and deletion) that exhibits timing unpredictability and that retains its efficiency despite adversarial attacks. To demonstrate these advantages, we develop a framework by which an adversary tracks a probability distribution on the data structure's state based on the timings it emitted, and infers invocations to meet his attack goals. © 2009 Springer Berlin Heidelberg.
CITATION STYLE
Bethea, D., & Reiter, M. K. (2009). Data structures with unpredictable timing. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5789 LNCS, pp. 456–471). https://doi.org/10.1007/978-3-642-04444-1_28
Mendeley helps you to discover research relevant for your work.