Introducing Connectivity

Why network?

We need to communicate from one machine to another, ranging from (for example) PC to printer through to mainframe to a user's terminal. Computers, printers, communication devices (such as gateways, bridges and servers) are networked together to allow the sharing of information, software and other resources.

There are two main reasons for using a network:-

1. information sharing

2. resource sharing.

Networking facilitates:

  • sharing of information (data)
  • sharing of peripherals (printers, plotters, fax, modems, scanners, tape streamers, CD-ROMs etc.)
  • sharing of software (if the license allows this)
  • easy implementation of back-up strategies for data and software.

It also means that workstations do not have to have a large hard-disk (or even a disk at all). It is important to note that software must be set to have a standard configuration protected from user modification.

Distributed processing
Current trends are moving towards 'distributed processing'. Here distributed processing refers to a method of data processing where files are stored in different locations and processing may take place via many different locations. Processing functions and data are distributed throughout the organisation according to when they are needed (and by whom). Newer software applications better facilitate distributed processing and allow users to work together in work-groups and electronically share vital (work-group relevant) resources.

Three types of network

Local Area Network (LAN)
A LAN can be considered as a bi-directional communications network that operates over a limited geographical area. The LAN typically facilitates the connection of a group of computers, terminals and other hardware over a limited geographical area, for example it may connect devices that are relatively near to each other (for example in one building). Devices can communicate with other devices using the same common physical transmission medium. In other words, a transmission from one device can be received by all other devices on the same network. A LAN is usually owned and controlled by one organisation, and typically is limited in size according to the organisation's premises. They can be linked to other LANs or linked to WANs. Low speed LANs might operate in the 1-20 Mbps, whereas high speed LANs might operate at speeds greater that 100 Mbps. LANs are a combination of hardware, cabling (and associated components), software and users brought together to form a coherent system. They facilitate the sharing of costly resources such as printers. In addition, 'distributed and Client Server Architectures can be employed further providing for the sharing of resources, as well as providing mechanisms to promote group working and sharing of data.

The LAN will consist of:

  • A physical medium whereby data may travel at a high transfer rate with a low possibility of errors.
  • A physical medium laid out in a particular topology. A topology is basically the pattern that you lay out the physical medium (wiring) for the network.
  • An access method whereby data are permitted to get onto the physical medium.
  • A transmission technique appropriate to the data that are transmitted - such as Baseband or Broadband.

Approaches to LAN networking include the following:

  • PC local network - this is mostly used for the sharing of applications and data within a room / department.
  • Backend network - this may be used to connect large computer systems together, for example two different mainframes run be the same company.
  • High speed office network - here we would consider higher specification devices and connection mechanisms to ensure speedy document and image transmission.


LAN Topology
All networks are classified into general categories according to their pattern, this is known as Topology - the main LAN topologies are Star, Ring and Bus. The topology basically describes the way in which devices share the transmission medium.

All devices are attached to a common central node (hub). A typical hub will accept data from the sending device and deliver the data to the intended receiving device.

All devices are connected in a closed loop - a cable connects the first device to the second device, and so on. The final device will be attached to the first device. The network may not necessarily be in the pattern of a 'real' ring, so the term ring applies to the logical connection between devices. Basically the network consists of a number of repeaters joined by point-to-point links. Each receiver receives a 'bit' and then passes it on. The links themselves can send data both directions.

All devices are attached to a single backbone - a cable capable of linear transmission. Any attached device may send a signal down the cable. All attached devices can hear and receive the signal. In order to regulate data flow, the data are sent in frames. Each frame will also contain data indicating both the source device ant the intended destination device.

Metropolitan Area Network (MAN)
The MAN will cover a slightly wider area than for example the LAN. It might for example include a city-wide area, for example across several buildings or a campus. The MAN will interconnect the resources available throughout a number of buildings. It requires moderate to high-speed data transmission rates.

The MAN may be used to provide means for internetworking several LANs. As such they may be a mixture of rented and owned, or owned by one orgamsation.


Wider Area Network (WAN)
The WAN will covers a wider area; for example buildings, cities and countries can be networked together. High bandwidths and subsequent transmission speeds are a must - as data may have to cross vast areas, cross international boundaries, and various regulatoiy bodies.

LANs and MANs can be used to create WANs. Their nodes (the points where devices area attached to network) are made available to a wider range of users spanning a larger geographical area. WANs typically use point-to-point lines with high speed capacity. Users may rent links into a WAN.


Baseband and Broadband
Baseband uses the entire available bandwidth, and so only allows one digital signal at a time to travel along the transmission medium (the cable). Information is represented by changes in the voltage. There may be more than one device using the same transmission line, provided there is a method for controlling access to the transmission medium. The data signal may be sent bi-directionally - regardless of the transmission start-point the data may be directed either way on the transmission medium. The signal itself is not be changed by modulation. Baseband can be employed using twisted pair or coaxial cable. Baseband uses bus topology and can be used to transmit data only a few kilometres. Baseband carries a signal such as a digital signal in the form of pulses placed directly on the bus. LANs using Baseband buses can typically facilitate transmission over distances up to 1km.

Broadband technology is more complex than Baseband, however it is able to carry more information and support a greater number of devices. (Available bandwidth will be greater than 4KHz - which is minimum required for a voice grade channel). Broadband can also be employed using coaxial cable. Data may only be sent unidirectionally. Analog signals are employed to transmit the data (in radio frequency mode) and so a RF modem is required. Using frequency division multiplexing (FDM) it is possible to have multiple data channels, to support transmission of various data types, including video and audio. Far greater propagation of signals is possible (depending on configuration). Data may only be transmitted one direction - therefore start-point data will have to move down-stream prior to getting to end-point. Broadband can employ bus or tree topology. Broadband may use an analog signal using 'frequency division multiplexing'. This will permit far greater transmission distances when compared with Baseband.

What are the main difference between Baseband and Broadband? Basically with Baseband each single wire will carry a voltage signal that will go on to represent one channel of information. With Broadband a single wire can carry a spectrum of carrier waves which can go on to represent a number of channels of information along the same single wire.

Transmitting data
Data are gathered into packets. The originating node will place the data onto the transmission medium -. the cable.
Transmission involves the following stages:

  • Each sending devices will broadcast a message on cable.
  • The recipient reads the message from the cable.
  • Other devices ignore message.
  • Data are passed along the transmission line.
  • All other nodes are aware that data are being transmitted.
  • Only the intended node is permitted to retrieve the data from the transmission line.

The message is simultaneously broadcast to all nodes (data are placed on the transmission line) on the network. All nodes can 'hear' the message. The receiving node will examine the data. The data will contain the address information
of the intended node. If the address matches the receiving node's address it will be accepted. The message itself has four main parts:-

  1. Destination address - A bridge will read the destination address from the data frame. The bridge will determine if the address is local or remote. Only data with non-local addresses will be permitted to cross the bridge.
  2. Source address - The bridge will also check the source address of each data frame received.
  3. Data
  4. Error check.


Medium Access Control methods

All LANs will consist of a number of devices that have to share the networks transmission medium. As such a mechanism is required to control the access to this transmission medium. The Medium Access Control (MAC) aims to regulate access in order to provide for efficient use of the transmission medium's capacity. There may be one device that will perform this controlling duty, or all of the devices may collectively perform this duty.
Any network must provide for a democratic medium - in other words all devices should be allowed the opportunity to send data along the transmission medium. Any device may attempt to send data at any time, and as such there must be controls in place to govern data transmission. This is especially important on a Baseband network where only one data signal is permitted to travel along the transmission line at any one time. As such, an access method is required to allow equal availability of all the network to each attached device. Two main access methods used in LANs include CSMA/CD and token passing.


This is the most commonly used protocol. Notably the original Baseband version is called the Ethernet, and the Broadband version is called MITREnet. Carrier Sense Multiple Access with Collision Detection. CSMA/CD listens to the network prior to transmitting - it tries to sense the presence of a carrier. If a carrier is sensed, this would imply that the transmission medium was in use. If no carrier was sensed, the system assumes that transmission can start. It continues to listen throughout the transmission, and will detect any collisions. If a collision between two sending devices is detected, both devices are told to stop transmitting immediately. Transmission begins again after a random back-off period. CSMA/CD is the access method used by Ethernet and other LANs.

The following describes the typical actions for any device wishing to send over a network that uses the CSMAICD protocol:

  1. Device listens
  2. If the transmission medium is idle then begin transmission (go to step 4.), else go to step 3.
  3. If the transmission medium is busy, continue to listen until it is idle, then begin transmission.
  4. If a' collision is detected during transmission send a brief jamming signal to advise all other devices that there has been a collision, then you stop your transmission, and return.
  5. After sending the jamming signal wait a random amount of time, and then re-send (step 1.)


Token passing
A token is a special kind of data packet (a small frame) that is passed from device to device, as in a relay race. The possession of the token permits the device to transmit data on the transmission medium. If there is no token there is no permission to transmit. When the device has finished transmitting it passes the token on to the next device. If that device does not wish to transmit it also passes the token on, or else the device begins transmitting. The IBM Token LAN is a major user of this access method.


PC/LAN connection
All devices on a LAN connect to a common cable - the transmission medium. The cable is usually twisted pair copper wire, coaxial cable or fibre optics. Two common types of connection are:

1. Ethernet (10 - basic or 100 Mbps - fast)
Ethernet is a LAN standard and uses Baseband transmission mode. It uses coaxial cable (or twisted pair wire), a bus topology and CSMA/CD. Baseband will use the entire available bandwidth for one channel - hence the reason why only one data signal can be sent at any one time.

2. Token Ring (16 Mbps)
A token ring is a LAN with a ring topology that uses 'token-passing' in order to gain access to the network.

A gateway is used to provide an information link between networks. It is a device that connects two dissimilar networks together. The gateway enables the networks from two different vendors to communicate. It will provide a transmission service from one vendor's protocol to another. In other words a gateway is used to translate from one WAN protocol to another, for example a TCP/IP network to an IBM SNA network.

The router acts as an traffic interface between networks. It primarily aims to connect two similar networks using the same protocols. It also' chooses the best 'route' between two networks where there are a number of paths. Routing decisions will be based on the network layer information and routing tables derived from related protocols. It is a communications interface and will seek to move packets of data over the established link.

Bridge (a router)
This forms a bridge between several individual smaller networks of the same type (with the same protocols), so resulting in a larger network. In other words a bridge is a device that connects smaller networks together on a local basis in order to exchange data. The bridge can be used to interconnect two or more LANs using the same or different LAN protocols such as token ring or Ethernet. There are several reasons for using bridges to connect together smaller LANs instead of building one large LAN, for example bridges may promote improved network: (1) reliability; (2)performance; and (3) security.

  1. One of the LANs may fail, however the other LAN can continue to operate;
  2. LAN performance decreases according to the load placed on the network, more smaller LANs allows for the sharing of transmission loads
  3. Different LANs can be tuned to accommodate different types of traffic and security needs;

Repeaters are used to regenerate signals. They are extremely useful when extending a LAN. Sometimes a Bridge can perform this function.

Other Comms Links

  • PSTN (Public Switched Telephone Network)
  • Lease/private lines (Kilostream, Megastream)
    The line will be leased by a private user. Services are subscribed to, and are for their exclusive and private use.
  • Microwave links
    This is high frequency transmission using microwaves to send high volumes of information. Individual telephone and telex circuits may be multiplexed for transmission over trunk routes by microwaves. (Multiplexing converts a number of different type signals into one type).


  • Modem (from words MOdulator & DEModulator)
    A modem is a data communications device used for the modulation and demodulation of data - that is changing analogue data to digital data and the reverse. Modems always have to be used in pairs and will be positioned at the sending and receiving nodes.
  • MegaStream Links
    A megastream link provides high speed digital conununications service on a point-to-point basis, and is offered by
    for example BT.
    - Rented, e.g.,
    - Pay fixed rate per month
    - No usage charge
    - Expensive
  • X25 Links
    X25 provides an interface between data terminal equipment and data circuit terminating equipment for devices operating in the packet mode on a public data network. X25 is an international packet switching standard. It can facilitate WAN links at speeds of up to 56 kbps
  • Nornial set-up is:-
    - X25 card in the server
    - Server connection to local provider
    - Client uses PSTN modem to access local PAD (packet assembler/deassembler)
    - Advantage is lower cost and better performance
  • ISDN (Integrated Services Digital Network)
    Computer and communication technologies continue to converge, as such there is a greater need for integrated systems that can transmit and process all types of data such as voice, image and video. ISDN seeks to provide for this. ISDN is intended to be a world-wide public teleconununications network that will eventually replace the existing public telephone network. As such, it hopes to deliver a wide range of services such as digital transmission and switching technology to support both voice and digital transmission. ISDN provides for a circuit-switched digital WAN service. It provides voice, data and image services in a unified manner. All data are synchronised so that all the data speak the same 'language' at the same speed. Basic concept is a dial-up digital line, where you pay for line rental and call charges

Advantages are:-
- Faster speed
- Better line quality
- Lower noise

Disadvantages include:
- Higher installation costs
- Higher line rental costs

Basic principles of ISDN:

  • Aims to support both voice and non-voice applications using a standardised set of facilities.
  • Supports both switched services and dedicated lines.
  • Uses up to 64kbps as its basic building block. This building block was originally chosen because of the rate at which digitised voice was commonly transferred. This is no longer the case for some newer standards.
  • Intelligence is built into the network - greater services now available, not just the telephone call!
  • It uses layered architecture which can be mapped into the 051 (Open Systems Interconnection) Model. Therefore standards already conforniing to 051 can be accommodated in ISDN.
  • A variety of configurations are possible.

This section has identified why there is a need to network a range of computer and communications hardware together. In particular, networking promotes sharing of information, peripherals and software. We have examined three types of network; Local Area Networks, Metropolitan Area Networks, and Wider Area Networks. We have examined connection methods, hardware and other communications methods used in networking. We have examined the physical layout of three types of networks, including the star, ring and bus topologies. Transmission methods have been explored, including Baseband and Broadband technologies. Finally we have looked further at a range of access and collision detection methods that can be implemented to support the successful transmission of data.