Network Topologies and Access Methods | ITL3 RS A1.2 |

Network Topologies and Access Methods

A network topology describes how computers and devices in a network are linked together.
Let’s take a look at some of the most common ones.

Bus Networks

Bus was mainly used in early networks. A single coaxial cable – the bus – connected each host device to the network. Terminators at each end reduced reflections – signals bouncing back along the network – that would cause interference.

Bus networks are not resilient as if one computer/host failed or a cable was damaged then the whole network went down.

Star Networks

In Star networks, each computer has its own connection to a central device. This could be a hub, switch, or router. Most networks today operate this way.

This is more resilient than a Bus network as the failure of a single computer or cable won’t knock out the rest.

Star-Bus Networks

Part of the Bus network idea is still around today. Two star networks are connected to make a star-bus network, although this isn’t the best use of available bandwidth.

Mesh Networks

In Mesh networks, every computer is connected to every other computer with no central connecting device.

Each computer requires a network card to communicate with every other computer in the network. (So, each computer in a network of four computers would need three network cards.)

Adding just one more computer would require an additional network card for each computer.

This way of working is highly resilient and is only used for critical applications where fault tolerance is an issue, and any down time would be damaging to the customer.

However, a fully meshed core network provides extremely high levels of availability and is used where connectivity is important to the customer.

A partial mesh is similar, but some of the connections are removed. All the devices are still connected to all the others but sometimes through one of the other core devices.

Ring Networks

With a ring network topology, each computer is connected to the others in a closed loop, typically with a coaxial cable. This is now an outdated concept in local networks.

Token ring technology uses this type of topology and a token passing system. Only the computer with the token can send information. The token is passed to the next computer to allow it to send its data.

Ring networks can be considered resilient if data can travel in both directions around the network.

A single failure in the network would still allow data to travel in the other direction.

Access Methods
The token passing system is an access method – a way to pass data from one device to another.

There are two more access methods to look at for now – we’ll be going into more detail later in your programme.

They are Carrier Sense Multiple Access with Collision Detection, and Carrier Sense Multiple Access with Collision Avoidance.

CSMA/CD - Carrier Sense Multiple Access with Collision Detection
Ethernet is a common LAN standard in networking today. It’s a group of technologies that define how devices communicate with each other.

Ethernet is standardised by the Institute of Electrical and Electronics Engineers (the I – triple E).

IEEE 802.3 defines the CSMA/CD standard.

Computers on an Ethernet network all share the same channel, so only one device can send a message at a time. If two or more messages are sent over the network at the same time, it is referred to as a collision and all messages are discarded before a retransmission is attempted again later with CSMA/CD, the network is a little like a party. Imagine you’re in a room with a group of friends. Everyone has something to say, and everyone needs to be heard clearly.

• So, you prepare what you’re going to say.
• Wait for a gap so you can speak.
• Start speaking.
• If somebody else starts to speak you stop.
• Keep what you want to say in your mind, ready for another go.
• Wait for another quiet moment.
• Start speaking again.
• And repeat the process until you’ve said what you want to say.

In networking terms, with CSMA/CD, each networking adapter monitors the network, uses timers to set waiting periods and buffers messages until they can be successfully transmitted. All this can happen in a period of 10 milliseconds (that’s 10 millionths of a second).

Carrier Sense Multiple Access with Collision Avoidance

In wireless Ethernet (Wi-Fi) networks, CSMA/CA – Carrier Sense Multiple Access with Collision Avoidance - is used, which is defined by IEEE 802.11.

Where collision detection re-transmits messages after a collision occurs, collision avoidance looks for a sufficient gap before sending a message. The network adapter listens to the Wi-Fi network to check for signals.

It’s a lot like crossing a busy road.

• Look left and look right to check for oncoming traffic.
• If you see traffic – wait.
• Once the road is clear you can cross.

The network adapter monitors the Wi-Fi and uses timers to check for a gap. It buffers messages until they can be successfully sent. Again - all in a few millionths of a second.

Collision avoidance is used here instead of collision detection as wi-fi networks reduce the power available for receiving while they’re sending – so they’re not able to detect collisions the way a wired network can.

We’ve looked at how networks can be connected in different ways – their topologies, along with some of the pros and cons for each.

Remember the most common topology is the star network, and a fully meshed network provides a significant amount of resilience.