Network Engineering Stack Exchange is a question and answer site for network engineers. It only takes a minute to sign up.
Sign up to join this community
I have read in many tutorials that Encapsulation is when the data go down the OSI model or the TCP/IP model and a header is added to it at each layer, and Decapsulation is when the data+headers go up the OSI model or the TCP/IP model and a header is removed at each layer.
But are Encapsulation and Decapsulation is really what happens inside an operating system, or are Encapsulation and Decapsulation just a theoratical concept that isn't really implemented in operating systems?
Encapsulation (prepending of lower-layer header) and decapsulation (removal of that header) happen within each layer.
The "payload" (SDU data) to send is passed from the higher layer, possibly along with some metadata. Then the layer creates its header, embeds (encapsulates) the payload to create its PDU and passes that down the stack to the next lower layer. The physical layer is an exception, it encodes the "payload" for the physical link and moves the encoded bits from one place to the other. When data is received all the steps above happen in reverse.
This answer shows the encapsulation in more detailed example.
Whether that happens within a host OS, inside a NIC, in a switch, router, ... is a matter of implementation. Traditionally, physical layer encoding and data link layer encapsulation happen inside a NIC, network and transport layer encapsulation in the OS's stack, application layer encapsulation in the actual application. Today, many functions have been "offloaded" down the layer stack, so a NIC can also be responsible for network and transport layer functions, including encapsulation.
