Genetic Algorithm for Base Station ON/OFF Optimization with Fast Coverage Estimation and Probability Scaling for Green Communications

Minimizing the power consumption while maximizing the quality of service has become a mainstream problem in green communications. The existing approaches of network configurations often ignore the computation complexity caused by the large number of user terminals. Our contributions mainly lie in two folds. We formulate an optimization problem to maximize the user terminal coverage ratio with a given number of activated base stations. We propose a novel genetic algorithm to optimize the ON/OFF status of base stations with fast coverage estimation, in which the scaling and selection operators are carefully designed to take the probability distribution of the estimated coverage ratio into account. Experiments have been conducted to prove the proposed algorithm for the network configuration for green communication.