Lesson 5 - Encapsulation and De-enapsulation Process
Lesson 5 - Encapsulation and De-enapsulation Process
Assuming that you have read my previous lesson about TCP/IP Layers, it's a good idea to take a closer look at the vertical transmissions between layers occurring on the sender host and the reverse process occurring at the receiving end.
So, the main topic of this lesson, is going to be encapsulation and de-encapsulation process.
NOTICE!
The terminology used in this and subsequent lessons is defined in Lesson 4.
Remember that TCP/IP model does not follow OSI model exactly. OSI model is now used as a reference. What's described in this blog is TCP/IP model, not OSI or IPX/SPX model. Look at the comparison below:
Application Layer ==> Transport Layer
All transmissions start at the application Layer (layer 5; keep in mind, that here we're talking about TCP/IP model NOT OSI model which uses 7 layers; just the answer to a comment I got). The application on the local computer (client) initiates a request destined for the application installed on some other remote host (server). That request CANNOT be sent directly on to the wire (layer 1). Instead, it must be processed by ALL the layers in between (layers 4 through 2) before it reaches the media. This way, the request from the application is sent down to the transport layer (layer 4) for processing. Application will choose one of the two common protocols used in the layer 4: either TCP, for reliable connections, or UDP for un-reliable connections.
Transport Layer ==> Internet Layer
The transport layer, upon receiving the request from the application layer, will process it according to what it has been designed to do, and is going to stick a so called header to the incoming data (for the details regarding all headers please, see the next lesson). This header will be understood and processed by the layer 4 at the receiving host (remote computer). This segment (as we call it a 'segment' now), is sent down to layer 3 for processing.
Internet Layer ==> Network Interface (Data-Link Layer)
The internet layer will process the incoming data from the layer 4 (which now consists of data or payload from layer 5 + layer 4 header) and will stick its own, layer 3 header, to the payload coming from the upper layers. Once it's done, it will send the packet (as we call it a 'packet' at this layer) down to the layer 2 for processing.
Data-Link Layer ==> Network Interface (Physical Layer)
The Layer 2, similarly to the previous layers (layer 4 and layer 3), is going to process the incoming packet according to the functions designed for this layer. Then, it will attach its own layer 2 header, and is going to send the frame (the name given to the layer 2 protocol data unit) down to the layer 1.
Physical Layer = Signaling
The layer 1 is going to transmit bits, 1s and 0s. Bits is the name given to protocol data unit at this layer according to the OSI model which is the reference for all models). Technically, Layer 1 is common for any networking model and follows IEEE specifications.
The process described above, attaching the controlling information in the forms of the headers, is called encapsulation.
Once the bits arrive at the destination computer (host), the reverse process takes place which is called de-encapsulation.
Network Interface (Physical Layer) ==> Data-Link Layer
The layer 1 is going to accept incoming bits and send them up to the layer 2 for processing.
Network Interface (Data-Link Layer) ==> Internet Layer
The layer 2 piece of software can properly interpret the header information (control information) initially attached by the sender's layer 2 process. So, it reads the layer 2 header, then strips this off, and the content of the frame (without the layer 2 header it's called a packet now), is sent up to the layer 3 for processing.
Internet Layer ==> Transport Layer
The layer 3 is going to perform similar actions that the layer 2 just did. It reads and processes the layer 3 header in the packet. Notice, that only layer 3 process understands the layer 3 header. Then, it removes the layer 3 header and sends the content of the packet (data without layer 2 and layer 3 headers is called a segment) up to the layer 4.
Transport Layer ==> Application Layer
I'm sure you have already guessed what is going to happen next. Yes, the transport layer will read the layer 4 header, which consists of the instructions what to do next. Then, uppon stripping off the layer 4 header it will send what was originally created (data request from the client software) to the appropriate process/application at the receiving host (server application).
When the server sends the reply back to the client, the whole encapsulation process will occur again. The client receives the reply from the server, and de-encapsulates the incoming data like explained above.
And this whole process is repeated back and forth until all data has been exchanged.
This process is presented with the below graphics for better understanding of this lesson.
In the next lesson, we will take a look at the content of those headers and go over and discuss few important fields. This will help us understand the basic communication process using TCP/IP model.
So, the main topic of this lesson, is going to be encapsulation and de-encapsulation process.
NOTICE!
The terminology used in this and subsequent lessons is defined in Lesson 4.
Remember that TCP/IP model does not follow OSI model exactly. OSI model is now used as a reference. What's described in this blog is TCP/IP model, not OSI or IPX/SPX model. Look at the comparison below:
Application Layer ==> Transport Layer
All transmissions start at the application Layer (layer 5; keep in mind, that here we're talking about TCP/IP model NOT OSI model which uses 7 layers; just the answer to a comment I got). The application on the local computer (client) initiates a request destined for the application installed on some other remote host (server). That request CANNOT be sent directly on to the wire (layer 1). Instead, it must be processed by ALL the layers in between (layers 4 through 2) before it reaches the media. This way, the request from the application is sent down to the transport layer (layer 4) for processing. Application will choose one of the two common protocols used in the layer 4: either TCP, for reliable connections, or UDP for un-reliable connections.
Transport Layer ==> Internet Layer
The transport layer, upon receiving the request from the application layer, will process it according to what it has been designed to do, and is going to stick a so called header to the incoming data (for the details regarding all headers please, see the next lesson). This header will be understood and processed by the layer 4 at the receiving host (remote computer). This segment (as we call it a 'segment' now), is sent down to layer 3 for processing.
Internet Layer ==> Network Interface (Data-Link Layer)
The internet layer will process the incoming data from the layer 4 (which now consists of data or payload from layer 5 + layer 4 header) and will stick its own, layer 3 header, to the payload coming from the upper layers. Once it's done, it will send the packet (as we call it a 'packet' at this layer) down to the layer 2 for processing.
Data-Link Layer ==> Network Interface (Physical Layer)
The Layer 2, similarly to the previous layers (layer 4 and layer 3), is going to process the incoming packet according to the functions designed for this layer. Then, it will attach its own layer 2 header, and is going to send the frame (the name given to the layer 2 protocol data unit) down to the layer 1.
Physical Layer = Signaling
The layer 1 is going to transmit bits, 1s and 0s. Bits is the name given to protocol data unit at this layer according to the OSI model which is the reference for all models). Technically, Layer 1 is common for any networking model and follows IEEE specifications.
The process described above, attaching the controlling information in the forms of the headers, is called encapsulation.
Once the bits arrive at the destination computer (host), the reverse process takes place which is called de-encapsulation.
Network Interface (Physical Layer) ==> Data-Link Layer
The layer 1 is going to accept incoming bits and send them up to the layer 2 for processing.
Network Interface (Data-Link Layer) ==> Internet Layer
The layer 2 piece of software can properly interpret the header information (control information) initially attached by the sender's layer 2 process. So, it reads the layer 2 header, then strips this off, and the content of the frame (without the layer 2 header it's called a packet now), is sent up to the layer 3 for processing.
Internet Layer ==> Transport Layer
The layer 3 is going to perform similar actions that the layer 2 just did. It reads and processes the layer 3 header in the packet. Notice, that only layer 3 process understands the layer 3 header. Then, it removes the layer 3 header and sends the content of the packet (data without layer 2 and layer 3 headers is called a segment) up to the layer 4.
Transport Layer ==> Application Layer
I'm sure you have already guessed what is going to happen next. Yes, the transport layer will read the layer 4 header, which consists of the instructions what to do next. Then, uppon stripping off the layer 4 header it will send what was originally created (data request from the client software) to the appropriate process/application at the receiving host (server application).
When the server sends the reply back to the client, the whole encapsulation process will occur again. The client receives the reply from the server, and de-encapsulates the incoming data like explained above.
And this whole process is repeated back and forth until all data has been exchanged.
This process is presented with the below graphics for better understanding of this lesson.
Click the picture to enlarge it.
In the next lesson, we will take a look at the content of those headers and go over and discuss few important fields. This will help us understand the basic communication process using TCP/IP model.
Comments
Post a Comment