the methods of traffic engineering’s optimization in case of data transfer by two routs

To prevent congestion in certain sections of multi-service networks, data flow management, which leads to a more proportional distribution of resources and improved network functioning, is implemented. The task of selecting routes for individual data streams (traffic class), taking into account requirements of QoS, is solved by the methods of traffic engineering. With the help of these methods, it is strived to load all the network resources maximally and balanced, so that the network, with a given level of service quality, has high total capacity as much as possible. However, at the present time, there are no rigorous well-founded solutions for the problem of distributing data flows between the selected routes, taking into account the possibility of additional control over their capacity. The article proposes a method for optimizing the distribution of data flows and the bandwidth of the channels (routes) used in each separate information direction, with different coordination of control actions at adjacent levels of the network architecture. We consider scientific and technical proposals for the implementation of the developed method as part of the mechanisms that implement the traffic engineering in modern multiservice networks. Purpose: increasing the efficiency of the multiservice network by optimizing the traffic engineering. Methodology in article analytical methods for solving optimization problems with a non-linear objective function and linear constraints are used. Results: analytical expressions were obtained for the optimal distribution of data flows and bandwidth of the used channels, which provides the minimum values of delay indicators, and the use of these expressions in calculations in the mechanisms of traffic engineering was suggested. Practical implications: the obtained results should be used in modern multiservice networks, which are implemented data flow management through the traffic engineering in order to improve the efficiency of these networks.