Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

On local CAC schemes for scalability of high-speed networks Javier Aracil, Senior Member IEEE, Jose A. Hernndez, Antonio J. Elizondo, Ral Duque, scar

Summary: On local CAC schemes for scalability of high-speed networks
Javier Aracil, Senior Member IEEE, Jose A. Hernández, Antonio J. Elizondo, Raúl Duque, Óscar
Gonzalez de Dios*
Universidad Autonoma de Madrid, EPS, C/ Fco. Tomás y Valiente, 11, 28049 Madrid
* Telefónica I+D, Emilio Vargas, 6, 28043 Madrid
Tel: +34 91497 2272, Fax: +34 91 497 2235, e-mail: javier.aracil@uam.es
Next generation networks are required to provide bandwidth on-demand for fine granularity sessions. In this
sense, centralized CAC (Connection Admission Control) approaches could suffer from scalability problems if
the number of requests for connections were very high. In this paper we investigate local CAC schemes where
the decisions are performed at the network edges, based on pre-calculated admission quotas.
Keywords: CAC, distributed control, high-speed networks.
Call admission control is an essential functionaly in networks supporting Quality of Service (QoS). In this paper,
we report about distributed CAC schemes that serve to alleviate the scalability and responsiveness problems of
their centralized counterpart. Actually, currently proposed Call Admission Control methods for most next
generation networks are centralized, thus complicating matters for scalability and robustness. On the other hand,
distributed CAC techniques show a better behaviour in terms of reliability. However, the CAC becomes more
sophisticated and difficult to implement and manage, primarily because resource brokerage is now carried out on
an individual basis in many different CAC units. Consequently, such CAC units must be either notified about the


Source: Rico, Javier Aracil - Departamento de Ingeniería Informática, Universidad Autonoma de Madrid


Collections: Computer Technologies and Information Sciences