HIERARCHICAL APPROACH TO BUILDING GENERATIVE NETWORKLOAD MODELS

Performance evaluation study of computer networks requires a concise d escription of the workload under which the performance is to be evalua ted. The performance evaluation of networks is an important field of s tudy today, because of the increasing usage of computer networks. In t he context of network sizing or tuning, it is often necessary to condu ct the performance evaluation studies under different load conditions. The repeatability of the experiments for different workload profiles, requires that the workload models generate the workload profiles para metrically. Such a model, should preferably be time-invariant, consist ent and generative. We propose a hierarchical approach for building ge nerative networkload (workload for networks) models, based on the Cont ext Free Grammar (CFG). We view the networkload as a sequence that can be generated from the rules of a CFG. Our approach combines the estab lished practice of viewing the workload as ''consisting of a hierarchy '' and the CFG description, to produce a generative networkload model. The time-invariance and generative nature are verified experimentally . The usefulness of the networkload model, in the study of a typical r esource management problem of a network, such as the optimal allocatio n of clients to servers, is illustrated by using the generative model as input descriptor to a queuing network model of a single server netw ork.