دانلود رایگان مقاله مدل چند منبعی برای تعیین ابعاد شبکه های داده

Abstract

Traffic modeling is key to the dimensioning of data networks. Usual models rely on the implicit assumption that each user generates data flows in series, one after the other, the ongoing flows sharing equitably the considered network link. We relax this assumption and consider the more realistic case where users may generate several data flows in parallel, these flows having to share the user’s access line as well. We qualify this model as multi-source since each user now behaves as an independent traffic source. Usual performance metrics like mean throughput and congestion rate must now be defined at user level rather than at flow level. We derive explicit expressions for these performance metrics under the assumption that flows share bandwidth according to balanced fairness. These results are compared with those obtained by simulation when max-min fairness is imposed, either at flow level or at user level.

6. Conclusion

We have proposed a new traffic model for evaluating user-level performance in data networks. The key characteristic of this model is to account for bandwidth sharing on the user’s access line. The results turn out to be very different from those obtained with usual models in practically interesting cases, like n = 100 users having different traffic profiles or access rates. They coincide only for large values of n, say n ≥ 1000. Simulations show that the results are approximately the same under flow-level max-min fairness. When max-min fairness is imposed at user level, the throughput performance of users with low traffic or low access rate tends to be better than that estimated by balanced fairness. One of the key benefits of the proposed multi-source model is to account precisely for the number of access lines n without the complexity of the finite-source model. For instance, traf- fic intensity (and thus link load) is an exogenous parameter of the multi-source model but an endogenous parameter of the finitesource model. Moreover, the normalization constant is explicit in the multi-source model, which greatly simplifies the computation of the performance metrics. A drawback of the multi-source model compared to the infinitesource model is the lack of a recursive formula for evaluating the normalization constant in the presence of a large number of different access rates. We let this for future work. Other interesting issues include the derivation of more accurate approximations in case fairness is imposed at user level and extensions of the model to non-elastic traffic (for instance, adaptive streaming traffic).