Application of the conditional gradient method to a network resource allocation problem with several classes of users

Abstract

© 2019 IOP Publishing Ltd. We consider a general problem of optimal allocation of a limited resource in a telecommunication network. The users are divided into several different groups (or classes), which correspond to different levels of service. The network manager has to satisfy users requirements and maximize the network profit, which gives a convex optimization problem with balance and capacity constraints. We suggest to solve this problem with a two-level iterative method where the conditional gradient method is used at its upper level. Then we can solve the lower level problem with simple decomposition algorithms. The computational experiments show rather satisfactory convergence of the suggested method.