This layer allows computing devices to exchange data across networks. It receives segments of data from the transport layer protocols and encapsulates them within packets. The packets contain address information identifying the source and destination devices, allowing packet switching devices such as routers to route them to the correct destination network.
There are two Internet layer protocols currently in use:
Internet Protocol v4 (IPv4): this is the most popular TCP/IP Internet layer protocol, which uses 32-bit addresses, superseding historically previous versions. The networking industry is slowly phasing out IPv4 due to the rapid growth of the Internet and the increasing demand for IP addresses, which it can no longer support. However, this will take considerable time as there are many systems and devices still dedicated to IPv4.
Internet Protocol v6 (IPv6): this is the replacement for IPv4. It uses a 128-bit address and is capable of supporting the expected future growth of devices connecting to the Internet.
The switchover from IPv4 to IPv6 was facilitated by the layered approach to network protocols. Because IPv6 can perform the same function as IPv4, there was no requirement to redesign the protocols above (TCP and UDP) or below the Internet layer.
This layer is responsible for preparing the data packets it receives from the Internet layer for transmission to the physical media connecting devices within the local network. There are three main types of physical media available:
Copper: coaxial, twisted pair.
Optical: single mode, multi-mode.
Wireless: WiFi, satellite.
Due the wide range of media, and supported technology, the network access layer is more complex than the other layers. Additionally, while the upper layer protocols within the TCP/IP suite are implemented in software, the network access layer must provide physicalconnectivity, thus it has both hardware and software components, typically implemented within a device’s NIC.
The primary functions of the network access layer are:
Accepting packets from IP and encapsulating them within frames. Different protocols can use different types of frames.
Converting the binary bits that make up the frame into a signal suitable for the type of media that is in use. For example, the bits are converted into an electrical signal for copper media and into pulses of light for optical media. Bits are converted to ultra high frequency radio waves on a wireless network.
Whereas the upper layer protocols are controlled by the agencies charged with maintaining the Internet (primarily the Internet Engineering Task Force), the sheer number of physical media available has led to many different protocols being designed and produced, often by commercial organisations.
Start of Figure
End of Figure
Over time, many of these protocols have become standardised and thus have become available for general use. The most commonly used network access Local Area Network (LAN) protocol is Ethernet and its derivatives.
The frames used by Ethernet totally encapsulate the IP packets sent from the network layer, so devices cannot directly read the IP addresses they contain. This makes it necessary for frames to carry their own source and destination addresses to ensure frames are delivered to the correct devices in the local network.
In an Ethernet network, this address is known as a Media Access Control (MAC) address.