The Improved Genetic Algorithm for Assingment Problems

In this paper, we describe a new mechanism of cellular selection as an improved genetic algorithm for some optimization problems like cellular channel assignment which have multi feasible/optimum solution per one case. Considering the problems and the nature of relationship among individuals in population, we use 2-dimension cellular automata in order to place the individuals onto its cells to make the locality and neighborhood on Hamming distance basis. This idea as 2D cellular automata Hamming GA has introduced locality in genetic algorithms and global knowledge for their selection process on cells of 2D cellular automata. The selection based on 2D cellular automata can ensure maintaining population diversity and fast convergence in the genetic search. The cellular selection of individuals is controlled based on the structure of cellular automata, to prevent the fast population diversity loss and improve the convergence performance during the genetic search.