Abstract: A method of allocating radio channels among tactical networks of radio transceivers which uses fuzzy logic to evaluate quality of communication and which uses a genetic algorithm to find an optimal channel allocation plan. The method generates plural plans which each have a random allocation of the radio channels to the networks, and evaluates each plan to provide a goodness vector for each of the plans. The goodness vector is a vector of goodness values which are continuous from best to worst. The process iteratively ranks the evaluated plans based on their goodness vectors, uses higher ranked plans to generate further plans which are also ranked, and deletes lower ranked plans until the top ranked plan or the goodness vector does not change for a predetermined period. The channels are allocated in accordance with the top ranked plan.