I had a glance on OSI layer and TCP/IP layer. But still confusion remains of not using two layers (presentation and session)
The OSI reference model is just that: a model. It specifies an "ideal" protocol stack. TCP/IP was developed before the OSI model and does not explicitly contain all the layers. The data link and physical layers are combined into one and everything above the network layer is regarded as an application.
The Presentation and Session layers from the OSI/ISO model are part of the Application layer in TCP/IP. In other words, in the TCP/IP model, the Application layer is responsible for all the functions that are performed by Presentation and Session layers in the OSI model. You may say that the OSI/ISO model is more detailed.
Between the two models is a fundamental difference. The reason for this lies in the fact that they are occurring at different times and with different missions. Example: presence of tos (qos) in IP -packet, which is a network layer. These mechanisms belong to the session layer. I can show you other examples of incompatibility between the two models. The OSI model however is detailed and accurate. Communications based on TCP/IP protocol stack are realized like OSI, to separate the services as sublayers. These elements I described more detailed in my book "Modern Computer Communications".
Best regards!
I think it would be better to say "These mechanisms belong to the session layer in the OSI model". As I said, the OSI model is just a model, it is not unique. TCP/IP makes different decisions, they are not wrong (or "inaccurate") just different.
That's what I mean. But the OSI model is not only a model. Based on it has conversion, eg. SS7... etc. I recommend you the book by prof. Tanenbaum "Computer Networks", ISBN: 0133072622. There you will find answers to many questions. But read the book slowly and carefully.
I disagree, the OSI is a model in the sense that you cannot buy an OSI protocol stack. Of course people implement a stack more or less along the lines of the OSI model.
By the way, I have been using Tannenbaum as a course book for 15 years.
The OSI v Internet technology and business model distinctions are historically interesting and significant, agreed. Both have roles and functions to play in both defining network abstractions in away other can understand, and specifying protocols for inter-operation of functional network capabilities, no doubt.
Given - present world - of heterogeneous networks combining network function virtualization/software defined network & hybrid cloud management/orchestration capabilities spanning the Internet of Things, I - admittedly not so humbly : ) - suggest it is past time for a new model.
Please see attached WiGiT v0.4 model, for the next - virtual, cross-network - wireless grid world for mobile clouds and the Internet of Things cognitively networked services - developed with support of National Science Foundation Division of Engineering, Industrial Innovation and Partnerships, Partnerships for Innovation Wireless Grid Innovation Testbed (WiGiT) project. and its over The Enterprise Cloud Leadership Council of TM Forum, IEEE and others contributed/offered feedback as well. Not quite ready for prime time as we are admittedly only at v0.3/v0.4.
I will be publishing several books in coming year which which further elaborate and which can permit all those willing to - create their own Heterogenous Networks for mobile clouds across wireless grids - do so readily, in sync with our model, and evolving open specs for wireless grid - networks.
Not the answer you were looking for but maybe one that helps?
- Badges
- Science topic
- Similar topics
- Computer Science