Effective solution for scalability and productivity improvement in fault-tolerant routing

Abstract

Effective solution for scalability and productivity improvement in fault-tolerant routing using the flow-based model was proposed. In practice, the faulttolerant routing based on reservation of network resources is usually associated with decrease in overall performance and scalability of protocol solutions. In this regard, in structure of the model introduced conditions that are responsible for the fact that the primary and backup routes differed only by network elements for which redundancy provided. This was reflected in modification of objective function within the formulation and solution of optimization problem of the fault-tolerant routing. Numerical examples showed efficiency of the proposed solution.

Enhanced method of fast re-routing with load balancing in software-defined networks

“…When solving the fast re-routing problem in SDN, we use the functional model described in [23][24][25]. It belongs to the class of flow-based solutions and focuses on the optimal use of available network resources during the process of routing in network.…”

“…The conditions (4) and (5) are linear in contrast to the nonlinear solutions proposed in [24], thereby reducing the computational complexity of producing the final protocol solutions.…”

“…In solving problems of single path routing (1), the following conditions for protecting the bandwidth of calculated paths are proposed in [24] …”

“…This is primarily because the calculation of the required routes is carried out on the SDN-controllers, mainly centralized by solving quite complex computationally optimization problems [13][14][15][16][17][18][19][20][21][22][23][24][25]. Implementation of the same principles of fault-tolerant routing is characterized by the fact that along with many primary routes it is necessary to calculate the backup paths using various schemes for protecting network elements: a link, a node or a path.…”

“…As shown by the analysis, on the one hand, the need to calculate a set of backup paths along with the primary routes implies a significant increase in the dimensionality of the formulated optimization problem [23][24][25]. On the other hand, the formalization of path protection schemes and their bandwidth also, as a rule, leads to a nonlinear formulation of the problem [24,25]. Therefore, the scientific task, connected with the development of effective computational methods for solving problems of fast re-routing with load balancing, aimed at increasing the scalability of their application in softwaredefined networks, is relevant.…”