By Raul Bernardino
Introduction:
Nowadays, the internet
technology and communication becomes a trend of the globalization in which
enabling us to communicate, to share information, and to search necessary
information from anywhere and at anytime with the no limitation. The
development of the networking technology brings internet communication throughout
the world, including reaching the very remote and isolated areas. Yet, the
exiting of the internet is to share knowledge and the best practices around the
world in order to improve the life of the communities and the society through text
files, emails, audio, and video streaming or in the real time interaction with the
audio conferencing and videoconferencing. Therefore, we have to manage
resources such networking, traffics, and internet bandwidth in order to meet
quality and best service delivery.
The data transmissions over
networking are sampling the data packets in to the frames and with best effort
of the internet protocols will send through communication media such switches
and router in order to reach the destination. The best efforts which are the
routers default here is still not guarantee that the deliveries are reaching
the proper destination within time frames. Therefore we have to overcome this
best efforts limitation such as packet lost and jitter including end to end
delays.
The packet loss is one
data-gram packet that loss during transmission from the source to the
destination because networks are overloaded. They (the packets) are retransmitted after
checked in destination packets. The retransmitted are not the case for the
multimedia.
Packet Jitter is one of
transmission datagram packets takes long time to reach the destination and also
have variation on the delays. Again this is not case for multimedia streaming
in which it has to be in fixed time frame.
End to end delay is one of
the delays that accumulate from prorogation delays that have happened in the
links, delay in queuing processes, and end system processes. The voice over internet protocols (VoIP) has
a certain fix delay in order to avoid the disturbing conversation.
The diagram below shows on
how data and voice are sent over the networks:
One reason why, we have to overcome the best
efforts as mentioned above is that the multimedia traffic requirement is
delivering packets in the fix time frames.
The real time conversation
for the “multiple
interactions the delays are acceptable between 150-400 milliseconds. Less than
150 milliseconds cannot hear by human listener. In other end the delays that
are more than 400 milliseconds will make frustrated,” Kurose
J. and Ross K., (2010, P. 627).
Therefore, the QoS comes to
play the rules. The QoS is standing for Quality of Service whereas an
architectural component that can be added in to the current infrastructure of
the IP. The architectural components as follows:
The classification of the packets: The nature of the applications
that transmitted over the networks needs certain type of the services.
Therefore we can classify each application based on the traffic on the network
and type services that needed. With the classifications of the applications,
the packets can be marked and allow the router to distinguish the priority of
the services for the delay sensitivities. For instance configure router with
new policy to treat packets accordingly as it shows in the below diagram:
The scheduling, policing, and isolation: Based on the class
marking of the each packet, we have done the isolation or distinguished the
packets among others. In other end it is also using schedule to determining
packet arrives in the network. For instance using FIFO, priorities, and round
Rubin scheduling as it shows in the below diagrams: FIFO Schedule:
Round
Rubin Schedule:
The high utilization of the resource: is to configure the router to reuse the allocation bandwidth
for certain classes that is not in use. This is to maximize the use of the
bandwidth as efficient and effective, as it shows in below diagram where it
allocates 1.Mbps for voice and video while data bandwidth allocation is 0.5
Mbps.
The admission call: The calls have to be declared
before it use. Because if it is exceeded the available bandwidth, it has to
block the call and give the busy signals. Therefore, it is needed to have a
flow declare before using the network, as it shows in the below diagram which
is one call at the time.
These are four components above are called four pillars.
In my own opinion scheduling, policing, and call admission can be combined in the QoS configuration. However, it is depending on how do we prioritize them, FIFO’s or round Rubin’s base.
References list:
1. Kurose, J.F. & Ross, K.W. (2010) Computer Networking: A Top-Down Approach.
5th ed. Boston:
Addison Wesley
2. University
of Liverpool/Laureate Online Education (2011) Lecture notes from Computer
Networking Module Seminar 6 [Online]. Available from: University of
Liverpool/Laureate Online Education VLE (Accessed: 9 September 2011)
3.
Bharadwaj, P (March 2005): Quality of Service in the Internet, [Online]. Available from: http://www.ias.ac.in/resonance/Mar2005/pdf/Mar2005p57-70.pdf
(Accessed: 9 September 2011)
No comments:
Post a Comment