Reliability prediction of LAN/WLAN integration network based on artificial intelligence

This paper describes the reliability and validation of prediction models of LAN/WLAN integration network. An improved PSO algorithm is used to optimize the weight of BP neural network. Support vector machine (SVM) is used in network reliability prediction. The LAN/WLAN integration network reliability prediction models are established with three methods (BP neural network, improved BP neural network based on PSO, and SVM regression model). The validity of reliability prediction model based on artificial intelligence as well as the advantage of using support vector regression method has also been demonstrated experimentally.