Abstract
The problem of adaptive routing in communication networks, i.e routing policy able to react to traffic variations, can be under quasi-static assumptions, formulated as non linear multiflow optimisation problems.Distributed routing is then equivalent in finding distributed algorithm solving the optimisation problem. Distribution means that decision (here computation) tasks are defined at each commutation node of the network, their coordination to achieve an overall optimum for the whole network being done trough information exchanges between the nodes. Different distributed algorithms based on primal or dual methods are presented and discussed with respect to their practical implementation
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have