Genetic Algorithms and Golomb Rulers
For my final year project in Computer Science I investigated the possibility of solving the Golomb Ruler problem using Genetic Algorithms. The paper researched the previous methods used in determining Golomb Rulers, and then built on a previous Genetic Algorithm based technique by using local search techniques to accelerate convergence.
The concept of a "Golomb Ruler" arose from the work of Professor Soloman W. Golomb of the University of Southern California. Golomb Rulers are a class of undirected graph measuring more discrete lengths than the number of marks they carry. In particular they have the property that on any given ruler, all differences between pairs of marks are unique. This property has elevated them from a mathematical curiosity to a powerful tool used in several different fields of research such as radio communications, X-Ray crystallography, coding theory and radio astronomy.
If you wish to know more, you can download the full paper below.
