Using Genetic Algorithms to Solve the Radio Link Frequency Assignment Problem

The Radio Link Frequency Assignment Problem (RLFAP) is that of assigning frequencies to a number of radio links in such a manner as to simultaneously satisfy a large number of constraints and use as few distinct frequencies as possible. This problem is known to be NP-complete. We describe the application of a genetic algorithm to the solution of the RLFAP. The standard crossover operators were found to be of limited use due to the highly epistatic nature of the problem and a range of new (crossover and mutation) operators are described. Dynamically modifying the weights used in the fitness function has also proved to be effective in improving the performance of the standard genetic algorithm. This work is being undertaken as part of the EUCLID CALMA Project - RTP 6.4, Combinatorial Algorithms for Military Applications.