Autocorrelation coefficient for the graph bipartitioning problem
Theoretical Computer Science
On the landscape ruggedness of the quadratic assignment problem
Theoretical Computer Science
SIAM Review
Fundamentals of Cellular Network Planning and Optimisation: 2G/2.5G/3G... Evolution to 4G
Fundamentals of Cellular Network Planning and Optimisation: 2G/2.5G/3G... Evolution to 4G
Partial neighborhoods of elementary landscapes
Proceedings of the 11th Annual conference on Genetic and evolutionary computation
Elementary landscapes of frequency assignment problems
Proceedings of the 12th annual conference on Genetic and evolutionary computation
Local search and the local structure of NP-complete problems
Operations Research Letters
Quasi-elementary landscapes and superpositions of elementary landscapes
LION'12 Proceedings of the 6th international conference on Learning and Intelligent Optimization
| Hi-index | 5.23 |
The Frequency Assignment Problem (FAP) is an important problem that arises in the design of radio networks, when a channel has to be assigned to each transceiver of the network. This problem is a generalization of the graph coloring problem. In this paper we study a general version of the FAP that can include adjacent frequency constraints. Using concepts from landscapes' theory, we prove that this general FAP can be expressed as a sum of two elementary landscapes. Further analysis also shows that some subclasses of the problem correspond to a single elementary landscape. This allows us to compute the kind of neighborhood information that is normally associated with elementary landscapes. We also provide a closed form formula for computing the autocorrelation coefficient for the general FAP, which can be useful as an a priori indicator of the performance of a local search method.