Service Model. The basic functionality of the service is the service customer’s ability to send IP packets to other cus- tomers, and to receive IP packets from other customers. It is a bare IP Internet connectivity service. The service is delivered to the customers by the Internet Service Providers. The basic property of the IP service model is that a Internet Service Provider (ISP) delivers the IP service to a set of customers. The most basic configuration of a ISP and two customers is shown below: Customer ISP Customer SLA SLA Figure 1. IP Service Model Each customer has its own Service Level Agreement with the ISP about the delivery of the IP service. The performance of the IP service is important for the customer, so the SLA contains a specification of what the minimum acceptable performance is. For this purpose, the SLA contains some parame- ters that are in effect boundary values for some particular IP performance measures. To be able to define these performance measures, a model of the IP service is needed that allows these measures to be defined. The most obvious place to look for such a model is the Internet Engineering Task Force [IETF], since this is the place where standardization activities regarding the Internet Protocol take place. The core protocol definitions were developed there [IPv4, IPv6], as are other main IP standards. Also for per- formance related issues there is work being done in the IETF, amongst others in the IP Performance Metrics working group [IPPM-wg]. However, the IETF has not defined a network model for IP on which performance measures are based. The ITU-T does have such a model, as specified in recommendation I.380 [I.380]. The model for an IP service and its associated performance parameters is constructed from the ground up there, and this model will be used as a basis for the work in this thesis. The terms used for the various com- ponents of the IP service are the same as are used there. An overview of the IP service model defined by ITU-T in recommendation I.380 is shown in Figure 2 below. Network Section Network Section Network Section gw gw gw SRC DST gw Router Gateway Router Source / Destination End System Measurement Point Figure 2. ITU-T IP Service Model The figure shows an IP infrastructure, and how it is built up from two basic component types: hosts and links. A host is a system that communicates using the IP protocol. If a hosts forwards incoming packets at the IP layer, and as such enables the communication between other hosts, it is called a router. A host where IP packets originate is called a source host, (marked SRC in the figure) and a host where IP packets finally arrive is called a destination host (marked DST in the figure). A link is a point-to-point connection for transporting IP packets between a pair of hosts. It is below the IP layer, so it does not contain any intermediate hosts. To allow for additional structuring of the IP infrastructure, network sections and circuit sections are defined. A network section is a set of hosts together with their interconnecting links that all fall under a single responsibility. A circuit section is a link that either connects a source or destination host to an adjacent host, or it connects a router in one network section to a router in another network section. Finally, the term gateway router is used to denote a router at the border of a network section that sends and receives packets across a circuit section to a gateway router in another network section. To facilitate the definition of performance metrics for the IP service, measurement points are defined. A measurement point is located at the boundary between a host and an adjacent link. Figure 2 shows two measurement points that allow performance metrics about the middle network section to be de- fined.
Appears in 2 contracts