Related Sites

NIC - Network Interface Cards

A network Interface Card (NIC) is the hardware device connected to your computer which provides the physical connection between your computer and the transmission media used to transfer data around your network. This could be Wireless, Ethernet, Token Ring etc...

NIC's can be connected to computers in a variety of ways. They can be part of the motherboard, where the NIC's components are actually built into the motherboard and the network port is usually found by the Parallel, Serial and USB ports at the back of the computer. The NIC could also be on an add on card which plugs in to one of the expansion slots on the motherboard (e.g. ISA, PCI or PCI-E slots). The NIC could also be a plug in device like a USB device or PCMCIA (PC Card) which connects to either the external USB ports or PC Card sockets on laptops. In general it does not really matter how they are connected to your PC although the connection type may have an effect on maximum attainable throughput (the speed of the connection).

NIC's are usually graded by speed, which is measured in bits per second (or more likely Mega bits per second Mbps). The main speeds that are supported by most wired network cards today are 10Mbps, 100Mbps and 1 Gbps (Giga bits per second). The main speeds that are supported today by Wireless cards are 11Mbps, 54Mbps or 248Mbps, however in general with Wireless connections the actual throughput can vary depending on wireless signal strength and local environment.

NIC's generally reside in the 2nd layer of the OSI Model.

Network Cable

Network Cabling is generally thought of in two types; Infrastructure cabling and Patch cabling. Infrastructure cabling is made from solid core cable and this is the cabling run around and in the walls of a building creating a link between the network ports and the patch panels, connecting the servers, networking devices and all the office computers together. Solid core cabling is used as it helps slightly reduce resistance and is easy to terminate in IDC connections. Patch cables are the cables used to connect the network ports to the actual equipment that they service, e.g. computers, routers, servers etc... Patch cables use stranded cable as this type of network cable is easier to manipulate and bend around.

When using Ethernet cables, which are the main networking cables / technology used today, you have 4 pairs of wires which are usually terminated by RJ45 sockets / connectors. These cables are usually terminated with the pins on the RJ45 connectors on each end of the cable corresponding with the same pin on the other end of the cable, e.g. the green wire going to pins 2 on both RJ45 connectors at each end of the cable. This is known as straight through cabling and is used to connect network devices (e.g. routers, switches) to computers and servers.

The other type of cable wiring method used is called the cross-over cable. This is where specific pins are reversed at one end of the cable. This method used to be used to connect two computers together with no networking equipment, i.e. one end of the network cable connected to one computers network port and the other end connected directly to the other computers network port, with no routers or switches involved. Nowadays most networking cards support Auto Crossover (usually noted as MDI / MDIX), which means that cross-over cables are not needed that much on newer equipment.

Ethernet network cable is graded in categories which, when used with the same category networking equipment, identifies the speed at which it can carry data. The main Categories of cabling used today are Cat 5, Cat 5e and Cat6 and the basic difference is the amount of twist per pair per meter. This means the green pair may have 5 twists per meter of cable, for example. The twists in the pairs help eliminate cross talk (or interference) between the pairs allowing for a higher throughput down the cable by being able to use higher frequencies. Although Cat 5 and Cat 5e were designed to accommodate 1Gbps networks they are usually used in 100Mbps networks and Cat 6 is used for 1 Gbps networks.

The other thing that needs to be considered when thinking about network cables is whether they are shielded or un-shielded cables. Shielded cables have a metallic or plasma based sheath around the length of the cable, on the inside of the outer jacket. This reduces the amount of interference which can get through to the twisted pairs. Most of the time un-shielded cables are used in networking equipment as when shielded cables are used all equipment connected to the shielded cables needs to be properly earthed or you do not get any benefit from the shield in the cable.

Whatever the category of cable the maximum cable length for the above mentioned categories is 100m. This means that the total length of the cable, which includes infrastructure and patch cabling should not exceed 100m in length.

Network Cable resides in the 1st layer of the OSI Model.

Hubs

Hubs are used to connect multiple hosts together. For example, if you have a group of computers that need to be networked together you would connect network cables to the network ports on the computers and the other end of the network cables would get connected to the network ports on the hub and this would provide a physical connection between the computers so that data sent from one computer to another has a direct path that it can travel down to get from the source host to the destination host.

The way that hubs work is that when a host sends data to the port on the hub, the hub replicates that data and sends it out of all of the other ports on the hub. If the source computer is connected to this hub the data will end up at it's desired location. This usually means that the bandwidth available to the hub, e.g. 100Mbps hub, will share this available bandwidth between it's ports, so if two pairs of host were communicating through the hub at the same time each pair would only see 50Mbps.

Hubs reside in the 1st layer of the OSI model.

Switches

A switch works similar to a hub in that data is sent to it from the source computer via the network cable to one of the ports on the switch.

The main difference between a hub and a switch is that a switch will inspect the headers of the IP datagram to look for the address of where the datagram is destined for. When the switch knows the destination address of the datagram, it checks a table held in it's memory to see if the destination address is connected to one of it's ports. If the address is connected to one of it's ports it forwards the datagram to that port only. If the destination address is not connected to one of it's ports the switch will send the datagram on to the next switch.

The main benefit of this is that you do not have a lot of un-needed data traveling around your network. The less traffic on the cables or airwaves of your network, the less packet collision there will be and the less re-transmitting of data will need to take place. Also, usually switches do not share bandwidth, they can open a full speed connection between two ports while maintaining another full speed connection between two different ports.

Switches reside in the 2nd layer of the OSI model.

Modems

Modem stands for Modulate / Demodulate. They are devices which are used to modulate an Analogue carrier signal (like a telephone connection) to carry digital data or demodulate an analogue signal to decode the digital information which has been transmitted over it. The way most people understand this is the binary numbers that are used by computers (the 1's and 0's) are modulated in to sound waves that can be transmitted over the telephone service.

Modems are usually categorized by the maximum speed at which they can transmit data and this is usually shown as Bits per second (bps). This maximum value is the value at which data can be transmitted when in best conditions, so a dodgy telephone line could reduce the actual throughput of the modem because packets which have picked up errors traveling down a poor quality line have to be retransmitted. Many methods have been implemented to increase speed like being made with in built error correction to again help to increase speeds and by reducing errors.

Modems reside in the 2nd layer of the OSI model.

Routers

Routers are pieces of networking equipment that are used to transfer packets from one network to another, known as routing. These are used on inter-networked computers so that data packets can be forwarded through the different networks until they reach the network where the destination host is located.

The easiest way to describe routing is to think of a block of streets. If you want to get to an address in your street you just travel down the street (or in computing terms packets are sent to the switch which then sends the packets to the relevant host located on the same switch). If, however you want to go to an address in a different street you need to go to a junction where the streets meet to change from one street to the next (again in computing terms, when sending packets to a different network the packets get sent to the switch which sends the packets to the router which moves the packets from one network on to the next).

The router knows where to send the packets destined for different networks as it holds a routing table in memory which tells the router the best route to take to get to the relevant networks incoming router. These routing tables are made of by the routing communicating with each other using routing protocols which work the best path to get to other networks.

Routers are sometimes combined to make devices like ADSL routers, which are ADSL modems with an inbuilt router or switch routers, where the router will have multiple ports on the internal network side.

As routers work in the 3rd layer of the OSI model they route packets according to IP Address, which enables large scaling of the inter-network.

Bridges

A bridge carries out similar functions as a router in that it separates 2 network segments and is used to transfer packets from one segment to the next. The main difference between a bridge and a router is that a router works in the 3rd layer of the OSI model and a bridge works in the 2nd layer of the OSI model.

what this basically means is that packets are routed using their MAC Address rather than IP Address. Due to this a Bridge is not able to distinguish networks as a router can by using the IP Address. This also means that by using Bridges in your network rather than routers your network will not be as scalable, however as bridges do not use the IP Address if you are moving your hosts around the network a lot you will not need to re-configure IP Addresses as a bridge will use the MAC Address which is associated with the network card and does not change.

Bridges reside in the 2nd layer of the OSI model.

Repeaters

A repeater is a networking device that can be used to increase the 100m limit of Ethernet cables. A repeater will receive the data in one port, which by this point will have weakened or degraded, amplify the signal and then place it back on the network cable through its other network port.

Repeaters are usually built into other devices like routers or switches so these devices are the point at which the 100m length will start from.

Repeaters reside in the 1st layer of the OSI model.

Gateways

A gateway is a host or device on a network which acts as an exit point from the network you are located on. Gateways are usually associated with routers, bridges or switches. When a default gateway is set on a host, if it needs to send data to a different network other than the one it is directly connected to it will direct the packet towards the default gateway, which will send the packets on their way.

Gateways are usually used to control private networks access to the outside world by also employing NAT, Proxy or Firewall.

Gateways can reside anywhere in the 4th to 7th layers of the OSI model.

The OSI Model

The Open Systems Interconnection (OSI) Model is a conceptual model which is based on the TCP / IP model. The OSI Model is split up into 7 layers which, when sending data, pass the packets from one level down to the next providing the relevant factors needed when sending data around a network.

These 7 layers are as follows: -

Layer 7 - Application Layer

Layer 6 - Presentation Layer

Layer 5 - Session Layer

Layer 4 - Transport Layer

Layer 3 - Network Layer

Layer 2 - Data Link Layer

Layer 1 - Physical Layer

The Application Layer is the layer of the OSI model which Applications directly access. Services are performed for the application in this layer and requests to the Presentation layer are issued.

The Presentation Layer is the layer of the OSI model which manipulates the data to provide a standard for the Application layer to work from. Encoding and encryption are carried out at this level.

The Session Layer is the layer of the OSI model which controls the session (or connection) between two hosts. It manages the connections and is responsible for establishing and terminating a session between hosts.

The Transport Layer is the layer of the OSI which is responsible for converting the data into TCP or UDP segments and the actual transmission of the data between the hosts. This layer is responsible for controlling the reliability of the connection through flow and error control.

The Network Layer is the layer of the OSI model which is responsible for providing a means of transferring data of differing lengths whilst still maintaining the quality of service requested by the Transport layer. It also provide a logical addressing scheme, e.g. Internet Protocol (IP), to make the segmentation of networks possible.

The Data Link Layer is the layer of the OSI model which provides the actual transmission of the data between hosts and manages the interaction of networking devices on a shared medium. It arranges the data received from the Network layer into frames, which are chunks of the data of a set size.

The Physical Layer is the layer of the OSI model which defines how the transmission medium is set up. This can include the specifications for voltage, cable specifications, pin assignments etc... This layer also carries out flow control, contention resolution and modulation, if needed.

The OSI Model - TCP / IP

The TCP / IP suite is based around a conceptual model consisting of only 4 layers. Although this may seem dramatically different to the OSI model, the different layers in the TCP / IP model carry out the same sort of tasks as those laid out in the OSI Model.

The four layers are as follows

Layer 4 - Application Layer

Layer 3 - Transport Layer

Layer 2 - Network Layer

Layer 1 - Link Layer

The Application layer consists of a set of services designed to provide a user interface for host to host communications. Applications will use these services to initiate different connection types between hosts.

The Transport Layer provides an end to end connection between two devices. The Application layer will use either the TCP or UDP transmission protocols from the transport layer which helps to provide a barrier to the Application layer from the actual process of transmitting the data.

The Network Layer is responsible for taking care of the routing part of the data transmission. It provides a single interface for the upper layers to access.

The Link Layer oversees MAC Addressing, defines the protocols for the actual transmission of the data and monitors the exchange of data between the host and the network.