Topics
features
Publications
Issue Archive
Algorithm That Synthesizes Other Algorithms for Hashing
 Created: Thursday, 01 July 2010
A synthesized algorithm is guaranteed to be executable in constant time.
An algorithm that includes a collection of several subalgorithms has been devised as a means of synthesizing still other algorithms (which could include computer code) that utilize hashing to determine whether an element (typically, a number or other datum) is a member of a set (typically, a list of numbers). Each subalgorithm synthesizes an algorithm (e.g., a block of code) that maps a static set of key hashes to a somewhat linear monotonically increasing sequence of integers. The goal in formulating this mapping is to cause the length of the sequence thus generated to be as close as practicable to the original length of the set and thus to minimize gaps between the elements.
The advantage of the approach embodied in this algorithm is that it completely avoids the traditional approach of hashkey lookups that involve either secondary hash generation and lookup or further searching of a hash table for a desired key in the event of collisions.
This algorithm guarantees that it will never be necessary to perform a search or to generate a secondary key in order to determine whether an element is a member of a set. This algorithm further guarantees that any algorithm that it synthesizes can be executed in constant
time. To enforce these guarantees, the subalgorithms are formulated to employ a set of techniques, each of which works very effectively covering a certain class of hashkey values. These subalgorithms are of two types, summarized as follows:
 Given a list of numbers, try to find one or more solutions in which, if each number is shifted to the right by a constant number of bits and then masked with a rotating mask that isolates a set of bits, a unique number is thereby generated. In a variant of the foregoing procedure, omit the masking. Try various combinations of shifting, masking, and/or offsets until the solutions are found. From the set of solutions, select the one that provides the greatest compression for the representation and is executable in the minimum amount of time.
 Given a list of numbers, try to find one or more solutions in which, if each number is compressed by use of the modulo function by some value, then a unique value is generated.
This work was done by Mark James for Caltech for NASA’s Jet Propulsion Laboratory. For more information, download the Technical Support Package (free white paper) at www.techbriefs.com/tsp under the Information Sciences category. NPO45175
This Brief includes a Technical Support Package (TSP).
Algorithm That Synthesizes Other Algorithms for Hashing (reference NPO45175) is currently available for download from the TSP library.
Please Login at the top of the page to download.
White Papers

