Using Genetic Programming to Improve the Performance of Wireless LAN Access Point Configuration

Radio communication speed has improved by leaps and bounds with developments in communication technology. In a wireless LAN, clients connect to a network to communicate with access points. Various approaches have benn used over the years to maximize the efficiency of access point configurations, such as the simplex method, Tabu search, and the genetic algorithm. This paper describes how genetic programming can be applied to the problem in a more deterministic function to improve the convergence speed. Linear genetic programming using arrays to shorten the running time and ”bonsai” manipulation, hierarchichal pruning of genes based on evaluation of terminal nodes, are proposed. An experiment was carried out to compare the performance with that of standard genetic programming. Consequently, it was confirmed that linear genetic programming reduced the running time and that the proposed ”bonsai” manipulation improves not only the convergence speed, but also the evaluation results.