Layers of Encapsulation: Exploring the TCP/IP Model in Internet Communication

Table of Links

Abstract & Introduction

Definition of Key Terms

Fundamentals of Internet Operation

Encapsulation

Usage-Based Economic Models

Net Neutrality

Legal History

Researcher Conclusions

Areas for Further Exploration & References

4. Encapsulation

All of this works through a process known as Encapsulation [Fig. 2]. All packets sent over the internet are encapsulated. In the TCP/IP Model, there are four layers of encapsulation on the Internet.

4.1. Network Access Layer

The Network Access Layer (sometimes called Media Access Control or MAC) is the lowest layer, or Layer 1 in the model. This layer allows network activity at the most basic level and only for a single local network. It works at the network card level.

4.2. Internet Layer

Layer 2 (RFC 791, Sec. 2.2.) of the TCP/IP stack, this layer controls network activity between networks and on a global scale. It defines things like IP addresses.

4.3. Transport Layer

Layer 3 of the IP controls the transfer of data. After the previous layers have done their jobs of determining the address where a given packet should arrive, the Transport layer does the job of actually transferring the data and confirming it was sent and received correctly.

4.4. Application Layer

Layer 4 of the IP is the Application Layer. This might be a web browser or email program, Facebook or Twitter, or it might be a wind sensor or a fitness app on a watch.

4.5. How It Works

When an application is ready to send anything over a network or the Internet, it need only pass that data to the next layer down. Big pieces are broken into smaller pieces as necessary and each one of those pieces is then encapsulated – assigned some unique information that will allow it to be reassembled by the corresponding layer at the other end. Note that encapsulation does not include encryption which is a separate process.

Now that the Application layer has encapsulated the smaller pieces, they can be handed off to the lower layers. Then the lower layers take care of their respective network duties, further encapsulating each packet at each layer. Until they are sent over the network and routed to their final destination.

The process is repeated in reverse at the other end. Each layer peels off the encapsulation and passes the data up to the next level until all the data has arrived at the Application Layer of the receiving host. Here, the Application layer reassembles all the smaller pieces back in to a complete image, movie, email or anything else.

This process means that an application programmer does not need to worry about the mechanics of the network layer, and the network engineers don’t care what language the applications programmer uses.

There is a guiding principal that each layer should operate independently of the others. There are, however, situations that require one layer to utilize information from another layer. While there is an active effort to minimize this, there are certain security features and other functionality that requires the layers to interact.

Any change of this magnitude requires years of research and consensus from the ISOC’s various advisory boards, specifically the Internet Engineering Task Force (IETF) via the Internet Advisory Board (IAB) and the Internet Research Task Force (IRTF), as well as input from many outside sources. One example of an upcoming change of this magnitude is QUIC – a possible new HTML standard (in the Application Layer) developed by engineers at Google, that is based on UDP instead of TCP (in the Transport Layer). All the discussion and development of any code or standards related to QUIC are available to the public to read/comment and/or contribute to in a public Github Repository[10].