Abstract

In multipath routing, maximization of the cardinality K of the disjoint-path set for a given source and destination assuming an upper bound on the differential delay D is one of the key factors enabling its practical applications. In the paper we study such an optimization problem for multipath routing involving maximization of K under the D constraint as the primary objective, and then minimization of the average end-to-end transfer delay for the fixed (maximum) K under the same D constraint. The optimization approach is iterative, based on solving an inner mixed-integer programming subproblem to minimize the delay for a given value of K and D. In order to increase the solution space, we consider the strategy of allowing controlled routing loops. Such a technique is implementable in software defined networks and optical networks. We present numerical results illustrating the gain achieved by using controlled loops in comparison with the traditional loop-free approach.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call