Data link layer, Ethernet and physical layer


In the previous chapters, we learnt about the application layer and transport layer port communication, we also looked at the network layer and IPv4 addressing for remote delivery of packets. In this chapter we look at the data link layer and discuss how it prepares packets for transport over different types of media and physical network components, we will also look at some of the Ethernet concepts and finally look at the physical layer.

Data link layer

The data link layer is the 2nd layer in the OSI model. its function is to primarily prepare packets for transmission over the physical media. It also serves to control the flow of data that will traverse the physical media. The Data Link PDU is the Frame. The two main functions performed at this layer include.

  • Framing of packets received from the network layer
  • Control of how data is handled by the physical media through MAC (Media Access Control) and detection of errors.

The devices at this layer are referred to as nodes.

Protocols and services

In the other layers, most of the protocols were defined by the RFCs. However, in the Data Link layer, the IEEE defines most of the protocols. We will discuss some of these protocols in chapters on switching.

Physical layer

The OSI model layers that we have discussed previously mainly focus on conversion of the data from the user to a form that can be transmitted over the physical media. The physical layer which is the 1st layer of the OSI model, is responsible for transmitting the data over the different types of physical media that may be present.

The frames that are received from the data link layer are converted into bits for transmission over the medium in this layer. Depending on the type of physical medium, the physical layer may represent the bits as either; light signals, electrical signals or waves for transmission over wireless media.

The transmission of this bits over the physical media depends on the following:

  • The type of medium and its connectors
  • The form of representation of the bits; either electrical, light or wave signals
  • The data encoding and the control information
  • The types of transmitters and receivers in the networks

As we have learnt earlier; by this stage in communication, the user data has undergone several processes; segmentation at the transport layer, packets in the network layer, encapsulation into frames at the data link layer and finally, the data is converted into one of the three forms that can be transmitted over the physical media; electrical, light signals or microwaves in the physical layer.

The three main forms of transmission media that we use in networks are:

  1. Copper cable
  2. Fiber
  3. Wireless

Just like the data link layer, the standards in the data link layer are not defined under RFC, however, other standards do exist.

Data carrying capacity in the physical layer

The various physical media offer transfer of data at different speeds. The data transfer rate can be measured using three metrics.

  1. Bandwidth – the capacity of the line. This is measured in bp/s (bits per second), kb/s kilobit per second and mb/s megabit per second.

NOTE: you should be careful not to confuse the bandwidth units. The symbol used to represent bandwidth is bps or b/s. the letter b is in small letters, this is not the same as the storage speed which is measured in Bytes per second (Bps) with a capital B.

  1. Throughput – this is the actual transferred data over a certain amount of time, in most cases it is usually less than the bandwidth.
  2. Goodput – the actual useable data that has been transferred over a certain period of time is known as goodput.


The standards at the layer 1 and 2 of the OSI model are defined as Ethernet standards. The different standards used in Ethernet define the different layer 1 and layer 2 protocols, however, the format of the frame does not change.

As we mentioned in the previous sections, the data link layer provides mechanisms for converting packets to frames while the physical layer converts frames to bits which are then transmitted over the physical media.

The hub

In the yester years, the main Ethernet standards in many Local area networks was the HUB. The hub used a technique known as the CSMA/CA (Carrier Sense Multiple Access/ Collision Avoidance) and CSMA/CD (Carrier Sense Multiple Access/ Collision Detection). This was implemented using either the bus or hub technologies.

Like the name suggests, the hub was at the heart of communication in a network segment and devices could only transmit data one at a time.

The HUB operates at layer 1, this means that when data is received by a hub, it floods this information to all the devices in the network. This is a major performance issue since there may be congestion, network failure among others.

NOTE: the area in which the Hub is the center of communication is refer to as a collision domain. This is because there is a high possibility of collisions.

The communication in a collision domain such as the one by a hub has the following characteristics.

  • Flooding – where frames are sent to all devices in connected to the hub.
  • Only one device can communicate at any instance
  • Communication is only one way
  • One collision domain

The switch

The problems associated with hubs, such as high failure rate due to collisions were a major hindrance to growth of networks. As a result, a new solution to combat this was introduced. The switch, changed communication in networks in many ways, each port on a switch acted as a single collision domain, therefore, the switch had as many collision domains as it had ports. This meant that the likelihood of collisions was reduced. The switch worked at layer 2 as opposed to layer one which meant that the switch could learn of the devices on its network and instead of flooding frames, unicast communication was possible.

Further development lead to bidirectional communication, instead of only one way communication as seen in hubs, also with switches many devices can communicate simultaneously.

The area where a switch is center of communication is known as a broadcast domain. In this type of communication, each port is its own collision domain therefore, there are as many collision domains as there are switch ports.

Some of the other characteristics of a switch include:

  • Communication at data-link layer instead of the physical layer
  • Full duplex communication
  • Broadcast domains instead of a single collision domain

NOTE: Some of the concepts of switching will be discussed at a later stage. These concepts are vital and are frequently examined in the CCNA composite exam and the ICND 1 and ICND 2 exams.


In this chapter, we have looked at the data link layer and how communication at layer 2 occurs. We have also looked at the physical layer and sending of data over the physical medium as bits. We concluded with the Ethernet and looked at communication over a switch and a hub. In the next chapter, we will do the basic configuration of a router and look at some of the CISCO IOS basics.