A new multiple objective dynamic routing method using implied costs

Abstract

There are advantages in considering the routing problem in integrated communication networks as a multiobjective shortest path problem, having in mind to grasp eventual conflicts and trade-offs among distinct objectives and quality of services (QoS) constraints. On the other hand the utilisation of dynamic routing methods in various types of networks is well known to have significant impact on network performance and cost, namely in overload and failure conditions. This paper presents the detailed formulation of a proposal of a multiple objective dynamic routing method (MODR) of periodic state dependent routing type, enabling to represent distinct QoS related metrics and requirements in a consistent manner. The MODR method present formulation is based on a multiple objective shortest path model with constraints and is prepared to use implied costs as one of the metrics. Alternative paths for each traffic flow are changed as a function of periodic updates of certain QoS related parameters estimated from real time measurements on the routes and trunks of the network. Such paths are computed by a specialised and efficient variant of a bi-objective shortest path constrained algorithm, developed for the MODR, enabling to incorporate flexible requirements on the QoS metrics. The architecture of the routing system is discussed together with the features of its main modules. An illustrative example of application of the MODR path calculation module to a circuit-switched type network using blocking probability and implied cost as metrics, is also presented, considering different overload conditions.