T-copula Based Multivariate Estimation of Distribution Algorithm and Its Application to Wireless Sensor Network
Estimation of distribution algorithms are a class of optimization algorithms based on probability distribution model. Copula is a powerful tool for multivariate probability analysis. In this paper, a new T-copula based multivariate estimation of distribution algorithm is proposed and applied to optimization of wireless sensor network. In the algorithm, T-copula is used to build probabilistic model of the selected solutions. To estimate joint distribution of the selected solutions, the correlation matrix of T-copula is firstly determined by estimating Kendalls tau and using the relationship of Kendalls tau and correlation matrix. After the correlation matrix is determined, the degrees of freedom of T-copula is estimated by using the maximum likelihood method, and the Monte Carte simulation is used to generate new individuals from the estimated probabilistic model. Then, the proposed algorithm is applied to the layout optimization of wireless sensor network by maximizing network coverage and lifetime. The relative experimental results show that the algorithm has better performance and is effective.
Estimation of distribution algorithm T-copula Kendalls tau Monte Carte simulation WirelessSensor Network
Ying Gao Xiao Hu Huiliang Liu Fufang Li Yuanyong Deng
Department of Computer Science and Technology, Guangzhou University Guangzhou, 510006, P.R. of China
国际会议
太原
英文
14-18
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）