Distributed pre-computation for a cryptanalytic time-memory trade-off

Show full item record

Redirect: RIT Scholars content from RIT Digital Media Library has moved from http://ritdml.rit.edu/handle/1850/7805 to RIT Scholar Works http://scholarworks.rit.edu/theses/3139, please update your feeds & links!
Title: Distributed pre-computation for a cryptanalytic time-memory trade-off
Author: Taber, Michael S.
Abstract: Cryptanalytic tables often play a critical role in decryption efforts for ciphers where the key is not known. Using a cryptanalytic table allows a time-memory tradeoff attack in which disk space or physical memory is traded for a shorter decryption time. For any N key cryptosystem, potential keys are generated and stored in a lookup table, thus reducing the time it takes to perform cryptanalysis of future keys and the space required to store them. The success rate of these lookup tables varies with the size of the key space, but can be calculated based on the number of keys and the length of the chains used within the table. The up-front cost of generating the tables is typically ignored when calculating cryptanalysis time, as the work is assumed to have already been performed. As computers move from 32 bit to 64 bit architectures and as key lengths increase, the time it takes to pre-compute these tables rises exponentially. In some cases, the pre-computation time can no longer be ignored because it becomes infeasible to pre-compute the tables due to the sheer size of the key space. This thesis focuses on parallel techniques for generating pre-computed cryptanalytic tables in a heterogeneous environment and presents a working parallel application that makes use of the Message Passing Interface (MPI). The parallel implementation is designed to divide the workload for pre-computing a single table across multiple heterogeneous nodes with minimal overhead incurred from message passing. The result is an increase in pre-computational speed that is close to that which can be achieved by adding the computational ability of all processors together.
Record URI: http://hdl.handle.net/1850/7805
Date: 2008-10

Files in this item

Files Size Format View
MTaberThesis10-2008.pdf 558.4Kb PDF View/Open

The following license files are associated with this item:

This item appears in the following Collection(s)

Show full item record

Search RIT DML

Advanced Search