History in Data Structures: Faster Data Structures Through History Independence and Adaptivity
Spremljeno u:
| Izdano u: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Glavni autor: | |
| Izdano: |
ProQuest Dissertations & Theses
|
| Teme: | |
| Online pristup: | Citation/Abstract Full Text - PDF |
| Oznake: |
Bez oznaka, Budi prvi tko označuje ovaj zapis!
|
| Sažetak: | A key feature of dynamic data structures is that they must continually respond to a sequence of update operations (e.g., insertions/deletions) and maintain their ability to answer queries correctly, ideally while minimizing the cost of executing these changes. This dissertation examines the different ways that data structures interact with this history of operations that they experience. We view this through two lenses: the role of the history in the algorithmic choices that the data structure makes, and the reflection of the history in the data structure’s memory contents.With the classical security notion of history independence, data structures obscure their history of operations in order to safeguard information from adversaries. By enforcing that the memory representation of the data structure is independent of this history, the data structure ensures that even by gaining access to its entire memory representation, a malicious actor cannot discern any additional information, e.g., the order in which operations were executed or whether data was stored in the structure and later deleted. This work enhances the study of this property in two ways. First, we expand the types of data structures that can be made history independent with efficient and natural constructions, particularly in external memory. Second, we use history independence in a novel way as an algorithmic tool to achieve improved upper bounds, whereas it had previously been employed almost entirely to provide security guarantees. Subsequently, we present a data structure which in contrast, exploits and adapts to the history in order to effectively predict future operations, resulting in further algorithmic improvements. |
|---|---|
| ISBN: | 9798297604988 |
| Izvor: | ProQuest Dissertations & Theses Global |