A macro-level distributed method for the efficient allocation of zones in the Internet

Much previous work has been done on distributed dynamic resource allocation of resources like nodes and connected links in networks. It is not easy to find ways to allocate resources and analyze the performance of the way due to the scalability and change of the network. In this paper, we discuss a novel macro-level, zone-based approach. Zone-based resource allocation mechanism (ZRAM) has not been discussed in the research literature. By zone we mean, one or more than one neighboring autonomous system (AS). A zone has a zone manager and some border routers; which can cooperate with each other by exchanging traffic information. The key challenges for zone-based resource allocation are to realize the scalability and robustness. In the ZRAM, the network traffic is considered to be routed from zones to zones, not nodes to nodes. Routes on zones are decided based on states of the zones. Three types of zone state, vacancy, available and no-vacancy are obtained from the simulation results which are described later in this paper.