QoE-Centric Adaptive Multipath Routing under R-Factor with Load Balancing in SDN

In the paper, the QoE-centric adaptive multipath routing tensor model aimed at provisioning the R-factor in software-defined networks has been presented. The flow-based routing model that is the base of the proposed approach takes into account possible packet losses caused by congestion of network elements and includes the conditions for realizing a multipath routing strategy, flow conservation, and link overload prevention. A tensor network description has been applied in order to construct the analytical form of conditions for ensuring the Quality of Experience in terms of the R-factor, as well as the expressions for calculating the average end-to-end delay and the probability of packet loss. Finally, the technological problem of adaptive routing has been reduced to solving the optimization problem of nonlinear programming that made it possible to ensure the adaptability feature of routing solutions when an increase in QoE requirements caused an increase in network resources needed. Using the proposed model is characterized by the load balancing within the multipath obtained for the network. As a result, a numerical study on a network fragment has been demonstrated, which is confirmed the proposed model adequacy and effectiveness in terms of load balancing in a case of increasing the network load.