Exploiting relaxation in local search for LABS

Exploiting relaxation in local search for LABS

0.00 Avg rating0 Votes
Article ID: iaor20084086
Country: Netherlands
Volume: 156
Issue: 1
Start Page Number: 129
End Page Number: 141
Publication Date: Dec 2007
Journal: Annals of Operations Research
Authors:
Keywords: programming: branch and bound
Abstract:

Branch-and-bound uses relaxation to prune search trees but sometimes scales poorly to large problems. Conversely, local search often scales well but may be unable to find optimal solutions. Both phenomena occur in the construction of low-autocorrelation binary sequences (LABS), a problem arising in communication engineering. This paper proposes a hybrid approach to optimization: using relaxation to prune local search spaces. An implementation gives very competitive results, showing the feasibility of the approach.

Reviews

Required fields are marked *. Your email address will not be published.