An Network Reconfiguration Strategy for Customer Requirements Optimization
Ming Zhang, Chenglong Gong, Yanhong Lu · 2009
The main challenge of network reconfiguration is in avoiding deadlock anomalies while keeping restrictions on packet injection and forwarding minimal. In this paper, we present a novel network reconfiguration strategy for customer requirements optimization. First, we propose dynamic reconfigurable mathematical model based on genetic algorithm, neural network tools. Second, we use genetic algorithms, neural networks and artificial immune theory tools to study the reconstruction scope and the number of unit selection algorithm, propose the network layer dynamic reconfigurable algorithm and the objective function which include a number of factors, Last, we use artificial intelligence, network monitoring and management theory focuses on the network resource management strategy, information resources prediction algorithm based on historical, neighbor nodes monitoring strategies, resource availability query algorithm, the network security management mechanism of reconstruction network, synchronous real-time network performance monitor strategy to achieve network reconstruction.