Compaired to broadcasts and Multicasts, a Unicast is very simple and one of the most common data transmissions in a network.

The Reason for Unicast

Well it’s pretty obvious why they came up with Unicasts, imagine trying to send data between 2 computers on a network, using broadcasts ! All you would get would be a very slow transfer and possibly a conjested network with low bandwidth availability.

Data transfers are almost all of the times, unicasts. You have the sender e.g a webserver and the receiver e.g a workstation. Data is transfered between these two hosts only, where as a broadcast or a multicast is destined either everyone or just a group of computers.

Media Access Control (MAC) addresses are talked about in various sections on the site, such as the OSI-Layer 2, Multicast, Broadcast and Unicast. We are going to analyse them in depth here so we can get a firm understanding of them since they are part of the fundamentals of networking.

MAC addresses are physical addresses, unlike IP addresses which are logical addresses. Logical addresses require you to load special drivers and protocols in order to be able to configure your network card/computer with an IP Address, whereas a MAC address doesn’t require any drivers whatsoever. The reason for this is that the MAC address is actually “burnt-in” into your network card’s memory chipset.

The Reason for MAC

Each computer on a network needs to be identified in some way. If you’re thinking of IP addresses, then you’re correct to some extent, because an IP address does identify one unique machine on a network, but that is not enough. Got you mixed up?

Check the diagram and explanation below to see why :

You see, the IP address of a machine exists on the 3rd Layer of the OSI model and, when a packet reaches the computer, it will travel from Layer 1 upwards, so we need to be able to identify the computer before Layer 3.

This is where the MAC address – Layer 2 comes into the picture. All machines on a network will listen for packets that have their MAC address in the destination field of the packet (they also listen for broadcasts and other stuff, but that’s analysed in other sections). The Physical Layer understands the electrical signals on the network and creates the frame which gets passed to the Datalink layer. If the packet is destined for the computer then the MAC address in the destination field of the packet will match, so it will accept it and pass it onto the Layer above (3) which, in turn, will check the network address of the packet (IP Address), to make sure it matches with the network address to which the computer has been configured.

Looking at a MAC

Let’s now have a look at a MAC address and see what it looks like! I have taken my workstations MAC address as an example:

When looking at a MAC address, you will always see it in HEX format. It is very rare that a MAC address is represented in Binary format because it is simply tooooo long as we will see futher on.

When a vendor, e.g Intel, creates network cards, they don’t just give them any MAC address they like, this would create a big confusion in identifying who created this network card and could possibly result in clashing with another MAC address from another vendor e.g D-link, who happened to choose the same MAC address for one of their network cards !

To make sure problems like this are not experienced, the IEEE group split the MAC address in half, and used the first half to identify the vendor, and the second half is for the vendor to allocate as serial numbers:

The Vendor code is specified by RFC – 1700. You might find a particular vendor having more than just one code; this is because of the wide range of products they might have. They just apply for more, as they need !

Keep in mind that even tho the MAC address is “burnt-in” to the network card’s memory, some vendors will allow you to download special programs to change the second half of the MAC address on the card. This is because the vendors actually reuse the same MAC addresses for their network cards because they create so many that they run out of numbers ! But at the same time, the chances of you buying two network cards which have the same MAC address are so small that it’s almost impossible !

Let’s start talking bits and bytes!

Now that we know what a MAC address looks like, we need to start analysing it. A MAC address of any network card is always the same length, that is, 6 Bytes long or 48 Bits long. If you’re scratching your head wondering where these figures came from, then just have a look at the picture below which makes it a bit easier to understand:

So that completes the discussion regarding MAC Addresses! I hope you have understood it all because it’s very important so you can expand your knowledge and truly understand what happens in a network !

Routable protocols enable the transmission of data between computers in different segments of a network. However, high volumes of certain kinds of network traffic can affect network efficiency because they slow down transmission speed. The amount of network traffic generated varies with the 3 types of data transmissions:

• Broadcast

• Multicast

• Unicast

We are going to have a look at each one of these data transmissions because it’s very important to know the type of traffic they generate, what they are used for and why they exist on the network.

Before we proceed, please note that understanding the OSI Model (especially Layer 2 and 3), Ethernet and the way a packet is structured is fundamental to understanding a broadcast, multicast or unicast.

Network topologies can take a bit of time to understand when you’re all new to this kind of cool stuff, but it’s very important to fully understand them as they are key elements to understanding and troubleshooting networks and will help you decide what actions to take when you’re faced with network problems.

I will try to be as simple as possible and give some examples you can relate to, so let’s get stuck right into this stuff !

The Stuff

There are two types of topologies: Physical and Logical. The physical topology of a network refers to the layout of cables, computers and other peripherals. Try to imagine yourself in a room with a small network, you can see network cables coming out of every computer that is part of the network, then those cables plug into a hub or switch. What you’re looking at is the physical topology of that network !

Logical topology is the method used to pass the information between the computers. In other words, looking at that same room, if you were to try to see how the network works with all the computers talking (think of the computers generating traffic and packets of data going everywhere on the network) you would be looking at the logical part of the network. The way the computers will be talking to each other and the direction of the traffic is controlled by the various protocols (like Ethernet) or, if you like, rules.

If we used token ring, then the physical topology would have to change to meet the requirements of the way the token ring protocol works (logically).

If it’s all still confusing, consider this: The physical topology describes the layout of the network, just like a map shows the layout of various roads, and the logical topology describes how the data is sent accross the network or how the cars are able to travel (the direction and speed) at every road on the map.

The most common types of physical topologies, which we are going to analyse, are: Bus, Hub/Star and Ring

The Physical Bus Topology

Bus topology is fairly old news and you probably won’t be seeing much of these around in any modern office or home.

With the Bus topology, all workstations are connect directly to the main backbone that carries the data. Traffic generated by any computer will travel across the backbone and be received by all workstations. This works well in a small network of 2-5 computers, but as the number of computers increases so will the network traffic and this can greatly decrease the performance and available bandwidth of your network.

As you can see in the above example, all computers are attached to a continuous cable which connects them in a straight line. The arrows clearly indicate that the packet generated by Node 1 is transmitted to all computers on the network, regardless the destination of this packet.

Also, because of the way the electrical signals are transmitted over this cable, its ends must be terminated by special terminators that work as “shock absorbers”, absorbing the signal so it won’t reflect back to where it came from. The value of 50Ohms has been selected after carefully taking in consideration all the electrical characteristics of the cable used, the voltage that the signal which runs through the cables, the maximum and minimum length of the bus and a few more.

If the bus (the long yellow cable) is damaged anywhere in its path, then it will most certainly cause the network to stop working or, at the very least, cause big communication problems between the workstations.

Thinnet – 10 Base2, also known as coax cable (Black in colour) and Thicknet – 10 Base 5 (Yellow in colour) is used in these type of topologies.

The Physical HUB or STAR Topology

The Star or Hub topology is one of the most common network topologies found in most offices and home networks. It has become very popular in contrast to the bus type (which we just spoke about), because of the cost and the ease of troubleshooting.

The advantage of the star topology is that if one computer on the star topology fails, then only the failed computer is unable to send or receive data. The remainder of the network functions normally.

The disadvantage of using this topology is that because each computer is connected to a central hub or switch, if this device fails, the entire network fails!

A classic example of this type of topology is the UTP (10 base T), which normaly has a blue colour. Personally I find it boring, so I decided to go out and get myself green, red and yellow colours

The Physical Ring Topology

In the ring topology, computers are connected on a single circle of cable. Unlike the bus topology, there are no terminated ends. The signals travel around the loop in one direction and pass through each computer, which acts as a repeater to boost the signal and send it to the next computer. On a larger scale, multiple LANs can be connected to each other in a ring topology by using Thicknet coaxial or fiber-optic cable.

The method by which the data is transmitted around the ring is called token passing. IBM’s token ring uses this method. A token is a special series of bits that contains control information. Possession of the token allows a network device to transmit data to the network. Each network has only one token.

The Physical Mesh Topology

In a mesh topology, each computer is connected to every other computer by a separate cable. This configuration provides redundant paths through the new work, so if one computer blows up, you don’t lose the network On a large scale, you can connect multiple LANs using mesh topology with leased telephone lines, Thicknet coaxial cable or fiber optic cable.

Again, the big advantage of this topology is its backup capabilities by providing multiple paths through the network.

The Physical Hybrid Topology

With the hybrid topology, two or more topologies are combined to form a complete network. For example, a hybrid topology could be the combination of a star and bus topology. These are also the most common in use.

Star-Bus

Star-Ring

A network is simply a group of two or more Personal Computers linked together. Many types of networks exist, but the most common types of networks are Local-Area Networks (LANs), and Wide-Area Networks (WANs).

In a LAN, computers are connected together within a “local” area (for example, an office or home). In a WAN, computers are further apart and are connected via telephone/communication lines, radio waves or other means of connection.

How are Networks Categorized?

Networks are usually classified using three properties: Topology, Protocol and Architecture.

Topology specifies the geometric arrangement of the network. Common topologies are a bus, ring and star.You can check out a figure showing the three common types of network topologies here.

Protocol specifies a common set of rules and signals the computers on the network use to communicate. Most networks use Ethernet, but some networks may use IBM’s Token Ring protocol. We recommend Ethernet for both home and office networking. For more information, please select the Ethernet link on the left.

Architecture refers to one of the two major types of network architecture: Peer-to-peer or client/server. In a Peer-to-Peer networking configuration, there is no server, and computers simply connect with each other in a workgroup to share files, printers and Internet access.

This is most commonly found in home configurations and is only practical for workgroups of a dozen or less computers. In a client/server network there is usually an NT Domain Controller, to which all of the computers log on. This server can provide various services, including centrally routed Internet Access, mail (including e-mail), file sharing and printer access, as well as ensuring security across the network. This is most commonly found in corporate configurations, where network security is essential.