Network Resource Management Investigation using a Distributed Simulator

Network resource management deals with protocols and networks capable of performing a reservation of the available resources in order to guarantee a certain Quality of Service (QoS). Examples of these technologies are Asynchronous Transfer Mode (ATM) and Multi-Protocol Label Switching (MPLS), which are usually used in core networks. An important objective of network providers is to obtain the maximum profit from their resources, hence there is a need for an efficient resource management. There are different techniques and approaches in resource management: centralised, distributed, hybrid, based on artificial intelligence techniques, etc. Investigation in this field is difficult, mainly because network research laboratories do not dispose of a large core network where they can investigate their approaches and algorithms. This paper presents a simple but flexible distributed simulator that supports a wide range of different experiments. It is based on an event-oriented simulation at a connection level (no packet or cell granularity). The distributed simulator is oriented to the simulation of large core networks and support different routing and admission control algorithms. It also presents a definition of the scenario where this simulator can be used, mainly in the context of Traffic Engineering, i.e. dynamic bandwidth management and fast restoration mechanisms. Some results in the dynamic bandwidth management area are presented.