Networking Basics: In a network, there are many complex communications devices which are involved in management. Each of these enables access and connectivity to the network. This includes hubs and switches, repeaters, routers, bridges, gateways & wireless devices. Some of these are closely related, for example, switches and bridges. A bridge is a fairly simple technology, which is not usually used anymore. It would separate different parts of a network, reducing the number of network broadcasts.
When a bridge is powered on in an Ethernet network, they learn the network’s topology by viewing and analysing the source address of incoming frames from all attached network segments. A switch is a complex collection of bridges, developed as networking technology advances. This switch increases the speed of communication as it creates unique mini networks, also known as circuits. These can be implemented on a LAN to do three things: Increase the speed of connection
Control larger systems requiring a core switch to manage all other switches Aid control and security of the network via managed switches that can be divided into VLAN’s Switches come in a range of specs, coming in between 4 and 96 ports. This switch can create a virtual circuit between each communicating device, increasing the bandwidth, security and reducing the amount of collisions. Gateway: A gateway is a point on the network which acts as an entrance to another. In terms of the internet, a stopping point can be either a gateway node or a host node.
Both internet user’s computers & computers which serve pages to users are host nodes. Computers which control traffic within your company’s network or at your local ISP are gateway nodes. A default gateway is a device which may be a route or switch, enabling your device to connect another network – such as the internet. Hubs: These are used in Ethernet networks. Segments which use hubs are often described as having star topology, in which the hub forms the wiring centre of the star. In a hub, a signal that is received at any port is transmitted again on all the other ports. A hub is essentially a repeater, which operates on bits.
Using this hub provides fault tolerance, because each networks device has its own connection, and if this fails, only that single device is affected. Making the network larger is also easier, as many additional devices can be added to the network using a single hub, which is often connected to a network backbone. Repeaters: As a signal travels along a transmission medium, there will be a loss of signal strength. A repeater is a network device which receives the signal on one port, regenerates it, and then retransmits it on all remaining ports. These extend the length of a network (not capacity) by connecting two of the networks segments.
They can’t be used to extend the network beyond its limitations, or to connect segments which use different network access methods. However, they can connect different types of media and link bridge segments with different rates of data. Routers: Routers play an important role in guiding network traffic between particular networks. In packet switched network, the router determines the next point to which a packet should be forwarded to its destination. For home and business computer users, with high-speed internet connections, a router can act as a hardware firewall, even if the home/business only contains a single computer.
Network Topology: In terms of communication networks, a topology is a description of the arrangement of a certain network, which includes its nodes and connecting lines. When defining the network geometry, there are two ways, the physical topology and the logical topology. When choosing a topology for a network, you have to consider how it will affect how devices on your network communicate. On top of this, cost, scalability, bandwidth capacity, ease of installation and ease of troubleshooting. Physical topology of a network is the definite layout of workstations.
There are many physical topologies, 4 of which will be described below: A bus topology (hence the name), looks much like a bus line going through a city. The devices are connected by a single cable, running throughout the network. The main cable segment has to end with a ‘terminator’, which engages the signal when reaching the end of the line. Without this terminator in place, the signal that represents the data would reach the end of the copper wire and bounce back – causing network errors. This topology has a range of different advantages and disadvantages.
Good things about it are; it is low cost to install, uses a lot less cable than other physical topologies, workstations can be added easily, and it works well for small networks (between 2 and 10 devices). Disadvantages of bus topology are; it’s uncommon, so equipment available is on a small scale, and more expensive to buy. If the cable breaks, then the network is down completely, as well as it being difficult to isolate where the problem originates. Also, when you add devices to the bus, all devices are suspended from using the network.
Ring Topology: As the name also suggests, ring topology is connected in the form of a ring or circle. It has no beginning or end that needs to be terminated. This allows every device to have an equal advantage accessing the media. In a ring network, a single cable is shared by all the devices, and the data travels in one direction like a merry-go-round. Each device waits its turn and then transmits. When the data reaches its destination, another device can transmit. The most common implementation of the ring is the Token Ring network.
Ring topology has many advantages, these are: The performance of the network can be predicted precisely based on the number of computers on the ring. Also, using a dual ring gives some fault tolerance. Some disadvantages of ring topology are: It’s no longer a recommended option for new LAN installations, though popular in MANs, a ring network also requires a lot more cable than bus networks (though less than in star networks). Similar to bus topology, a break in the cable will bring the network down (unless dual ring), and again the same as bus topology if you add devices to the ring, all devices are suspended from using the network.
Star Topology: Another type of topology is star topology, this is, as the name again suggests, in the shape of a star. This topology is made of a connection point in the centre and a hub/switch where the cables meet. Each of the devices in this network is connected to the hub/switch with a separate cable. This requires more media, but has lots of advantages over both bus & ring topologies, which is why they have quickly become the predominant topology type for most of the networks.
These advantages are: You can easily add devices to your network as it expands, and the failure of a cable will not cause the network to be brought down. It’s also easy to find problems within the device and cable, and it’s the most common of the topologies, so equipment is readily available. Though it holds many advantages, it does have disadvantages, these are: This network requires more media than either of the two previous networks, the failure of the central switch can bring the entire network down, and the cost of installation and any extra
equipment is higher than the two previous networks listed. Tree Topology: This type of topology is like a mixture between bus and star topology. The structure (which looks vaguely like a tree) allows you to have many servers on the network and you can branch out the network in many ways. For colleges & universities this is helpful, so each of the branches can identify the relevant systems in their own network and connect to the big network in some way. Like the other topologies, this has both advantages and disadvantages.
The advantages are: Tree topology is supported by many network vendors, devices can be added easily, all computers have access to the larger and immediate networks, and it is the best topology for branched out networks. Disadvantages are: Maintenance of this network can be an issue when it spans a great area, and if the backbone fails, the entire network also fails. Network and Access Services: In network access, the most common method in use on LAN’s is CSMA/CD. This stands for Carrier-Sense Multiple Access/Collision Detection and can be broken up into three parts.
These are: Multiple Access – Meaning all systems can access it at any time, making it a contention method of control Carrier –Sense – This means the Network Interface Card on each of the systems senses whether or not there is traffic on the cable before sending Collision Detection – This means that collisions can happen, but only if two computers on the network send data at the same time – but the Network Interface Cards of the sending computers will detect that a collision has occurred so they can re-send their data.
This method holds many advantages and disadvantages, the advantages are as follows: It is reliable, as any collisions are detected and the packets involved are re-sent, preventing data loss. It is also relatively fast, with computers not having to queue to send/receive data. The disadvantages are, that it has limited range, and is inappropriate for large networks. Token Passing: This is a network access method particular to token-ring networks. These are a form of LAN developed and supported by IBM, running typically at 4. 16 or 16 Mbps over a ring topology.
In this type of network, access is methodical as it is based on the possession of a small frame called a token. Having a token gives a node the right to talk. In order to talk, a node captures a token as it moves around the ring, inserts its data, target node address and control information into the token frame. It then changes this frame to a data frame, transmitting the frame around the token ring towards the destination node. Nodes not targeted by the data frame just pass the token on. The target node opens the data frame and reads the data stored inside.
After processing this data, the target node releases the data frame, and after a complete circuit, it finds itself back at its sending node. Section 2 – Communication Protocols All communications between devices require agreement of the format of the data. The set of rules defining a format is a protocol. At the very least, a communications protocol must define: Rate of transmission Whether transmission is synchronous or asynchronous Whether data is transmitted in half-duplex or full-duplex mode Protocols can also include sophisticated techniques for detecting and recovering some transmission errors and for encoding/decoding certain data.
There are many properties of a transmission that a protocol can define. Common ones include: packet size, transmission speed, error correction types, address mapping, flow control, packet sequence controls, routing, and address formatting. Popular protocols include: File transfer protocol (FTP), TCP/IP, user datagram protocol (UDP), hypertext transfer protocol (HTTP), post office protocol (POP3), internet message access protocol (IMAP), simple mail transfer protocol (SMTP). TCP/IP Protocol Suite – This is the most widely used of the communication protocols.
A protocol suite consists of a layered architecture where each later has some functionality which can be carried out by a protocol. Each later usually has more than one protocol options to carry out the responsibility that the layer adheres to. TCP/IP is usually considered to be a 4 layer system, which are as follows: Application layer Transport layer Network layer Data link layer OSI Model – This is a conceptual model which characterises and standardises the internal functions of a communication system, partitioning it into abstraction layers. The model group’s similar communication functions into one of seven logical layers.
A layer serves the layer above it and is served by the layer below it. For example, a layer which provides error-free communications across a network provides the path needed by applications above it, while it calls the next lower layer to send and receive packets that make up the contents of that path. The model is split into 7 layers which are as follows: Application Presentation Session Transport Network Data Link Physical Error and Data Correction (P5) Here it will be explained how errors in transmission can be detected and what techniques can be used to reduce them.
In error and data correction there are many different types, I will be explaining and describing parity and modulus 11 checking. Data can be represented electronically, in ‘bits’ – a bit is a shortened version of the words ‘binary digit’ – these are represented using 0’s and 1’s. When bits are grouped together they can form more useful pieces of data, for example, 8 bits grouped together forms a ‘byte’. In networking, this data must be formed in a package to transport across the network; these are called ‘data packets’. This packet must at least contain the data and error control, as well as the source and destination address.
There are two types of transmission, these are synchronous and asynchronous – asynchronous means that the receiver must recognise receipt of the data before the sender sends more. In synchronous, both devices synchronise with each other before the data is sent. In terms of speed of data transmission, we can look at the bandwidth. This means the amount of data that can be sent through data transmission over a period of time. This is measured in bps, kbps, mbps and gbps (in ascending order of speed). The earliest error detection systems used parity bits, which is an extra bit added to the end of the data.
So all the data is in the first 8 bits and the ninth bit contains the parity data. The parity bit always has to be the correct number to make the data even, so if there are 4 1’s in the data, the parity bit has to be a 0. This type of detection is poor, with there being no way of seeing where in the byte that the error occurred. Checksums – This is a simple check that is used to detect errors in data. In the simplest form, a checksum is made by calculating binary values in a packet or another block of data using an algorithm and storing results within the piece of data. Checksums are very fast, with a high detection rate of 99.
6% errors in a packet. This level of reliability is suitable for simple communication situations, but less reliable than CRC (cyclical redundancy checking), which boast an even higher accuracy of over 99. 9% The receiving party relates the checksums calculated by the sender and by the receiver. If matching, the receiver accepts the transmission is free from errors. However, if they don’t match, we know there was an error. Error Correction – Cyclic Redundancy Check CRC is a method of error checking and correcting for data errors that have been transmitted on a communications link.
It was designed to detect accidental changes to the raw computer data, frequently used in storage devices such as hard disk drives. A CRC calculates the data using a binary sequence, returning a figure which identifies the data. This figure is called a checksum. When a block of data is received or seen, this sequence occurs, if the new CRC code is different to the one calculated previously, the block contains an error and action can be taken. These codes are unique error correction codes which can detect errors in data even when there is more than one altered bit.