Abstract
The pervasive use of the Internet, the world’s most extensive public communication system, for services ranging from academic research and e-mail to electronic commerce, necessitates that a policy should be put in place to ensure that such services are delivered in an efficient manner. Quality of service (QoS) is the capability of a network to provide better service to selected network traffic over various technologies. In this paper we focus on a specific aspect of QoS–namely, the path selection policy.
Key Terms in this Chapter
Preference Function: A preference function f has a domain set D consisting of multiple criteria x = (x1, x2, ..., xm), where xj represents the jth criterion. Each alternative is a point q = (q1, q2, ..., qk) in the set D. Building a preference function involves assessment of numerical values of the coordinates (q1, q2, ..., qk) for each alternative. Given x, y, z ? D then the following must hold: 1. x “is preferred to” y ? f(x) = f(y) 2. x– z “is preferred to” y – z ? f(x) – f(z) = f(y) – f(z)
Policy Based Metric: A characteristic of a network link that controls what type of traffic and how that traffic moves across the link. Policy metrics can be used to implement access control and data security of the link.
Concave Metrics: w is said to be a concave metric if w(P) = min(w(Li)). Example is bandwidth.
QoS Routing Algorithm: An algorithm that finds a path that meets a set of user constraints.
Strict User Constraints: Let S be a set of paths, then Lj are strict user constraints if: Lj = wj(P*) + ej, j = 1, ..., m where P* is the path for which 1 max( ( )) j m j w S = = is minimum and ej are small positive numbers relative to wj(P*). Strict user constraints imply that few paths exist that can meet the demands of the user.
Additive Metrics: Let w(P) be the total value for metric w on path P, and let w(Li) represent the weight of each link with respect to w on path P. Then w is said to be an additive metric if 1 ( ) ( ) li i w P = w L = S , where l is the number of links in the path. Examples are delay and jitter.
Optimization Goal: The user’s desire to find a path that minimizes or maximizes the value.