Research of Neural Network Architectures for Solving Adaptive Routing Problems in Multiprovider Networks of Distributed Data Centers

Abstract

Solving the adaptive routing problem in multiprovider networks of distributed data centers (DC) is a rather difficult task. The computational process of constructing routing tables is complicated with an increase in the load on network equipment and the need for its constant updating. The presence of several communication providers in the DC, the growth of user requests for various types of traffic also affect the choice of network construction option. The aim of this work is to implement the adaptive routing algorithm in multiprovider DC networks using a neural network theory. In the work a comparative analysis and experimental research of various architectures of neural networks were carried out and a model was developed that allows to find the optimal route in multiprovider DC networks with the highest accuracy.