Image Processing Reference
In-Depth Information
8.3 QoS Provisioning in Packet Switched Network
The IP network convergence [ 76 ] has imposed the Internet as being the core
infrastructure to attach various access technologies and to assure the transport of
all kind of services and applications (e.g., data, audio and video). A major challenge
in the convergence environment is that each service or application has its own QoS
requirements for bandwidth, latency, jitter and loss. For example, applications like
Voice over IP (VoIP) require 150 ms of (mouth-to-ear) delay, 30 ms of jitter and no
more than 1 % packet loss [ 77 ]. The interactive video or video conferencing streams
embed voice call, and thus have the same service level requirements as VoIP. In
contrast, the streaming video services, also known as video on-demand, have less
stringent requirements than the VoIP due to buffering techniques usually built in the
applications. Other services such as File Transfer Protocol (FTP) and e-mail are
relatively non-interactive and delay-insensitive. This means that the Internet must
be able to treat each service according to the service requirements and efficiently
utilize the network resources for operators to maximize revenue. However, the
legacy Internet system was not designed for these purposes; it treats all the services
equally in a best-effort fashion. In order to make the Internet more attractive, the
QoS provisioning consists of designing algorithms and protocols to enable the
network to provide predictable, measurable and differentiated levels of quality
guarantees. This is usually enabled through service admission control, service
prioritization and network resource reservation. The resource reservation, usually
resorting to the Resource Reservation Protocol (RSVP) [ 11 ], allows reserving
appropriate amount of bandwidth on the path to be taken by the packets that belong
to the service requiring the bandwidth. Further details are provided in the subse-
quent subsection.
8.3.1 Related Work
Network resource reservation control has become one of the most exciting frontiers
of research and development in the communications systems to deliver guaranteed
QoS. Key relevant standards of the QoS approaches include the IP Multimedia
Subsystem (IMS) from the 3GPP, the Resource Admission Control Function
(RACF) from the ITU-T (International Telecommunication Union Standardization
Sector) and the Resource and Admission Control Sub-system (RACS) from the
TISPAN (Telecommunications and Internet converged Services and Protocols for
Advanced Networking) [ 13 ]. Traditionally, the resource reservation is performed
on per-flow basis, meaning that the QoS signaling messages are triggered upon
every service request [ 10 ]. As a consequence, the approach has been criticized in
the research community due to the lack of scalability and therefore the energy
saving issues [ 12 ]. To ease the understanding, Fig. 8.8 illustrates the processes
related to QoS-enabled path setup for establishing a new service session, the path
Search WWH ::




Custom Search