Numerical Simulation and Experiments on Advanced Traffic Engineering

In this paper, we present artificial intelligence (AI)-assisted advanced traffic engineering (TE) methods and some techniques for numerical simulation and experiments to validate those methods. TE is to optimize network performance and resource usage by changing routes or logical topologies in response to environmental changes such as a sudden traffic demand increase. In performing TE, we modify the logical configuration of a network, but this often causes unexpected congestion or performance degradation. We cannot fully predict the behavior of an operational network composed of numerous heterogeneous network systems. To minimize the negative impact of TE, we develop numerical simulation and experiment techniques that emulate large-scale multilayer transport networks.