Chapter 2

• ARPAnet


  Advanced Research Projects Agency Network.
• WAN


  wide area network (network covering a very large geographical area).
• LAN


  local area network (network covering a small area such as a single building).
• MAN


  metropolitan area network (network whichis larger than a LAN but smaller than a WAN, which can cover several buildings in a single city, such as a university campus).
• Wide area networks (WANs)


  Wide area networks (WANs) are used when computers or networks are situateda long distance from each other (for example, they may be in different cities or on different continents). If a number of LANs are joined together using a router or modem, they can form a WAN. The network of automated teller machines (ATMs) used by banks is one of the most common examples of the use of a WAN.
• why do WANs usually make use of public communications network?

  Because of the long distances between devices, WANs usually make use of a public communications network (such as telephone lines or satellites), but they can use dedicated or leased communication lines which can be less expensive and more secure (less risk of hacking, for example).
• what will a typical WAN consist of?

  A typical WAN will consist of end systems and intermediate systems, as shown in Figure 2.3. 1, 3, 7 and 10 are known as end systems, and the remainder are known as intermediate systems. The distance between each system can be considerable, especially if the WAN is run by a multi-national company.

  
• The following is used as a guide for deciding the 'size' of a network:

  WAN: 100km to over 1000kmMAN: 1km to 100kmLAN: 10m to 1000mPAN: 1m to 10m (this is not a commonly used term – it meanspersonal area network; in other words, a home system)
• what are local area networks (LANs)?

  LANs are usually contained within one building, or within a small geographical area. A typical LAN consists of a number of computers and devices (such as printers) connected tohubsorswitches. One of the hubs or switches is usually connected to arouterand/ormodemto allow the LAN to connect to the internet or become part of a wide area network (WAN).
• what is metropolitan area network or (MAN)?

  - MANs are larger than LANs as they can connect together many small computer networks (e.g LANs) housed in different buildings within a city (for example, a university campus). MANs are restricted in their size geographically to, for example, a single city.

  - In contrast, WANs can cover a much larger geographical area, such as a country or a continent. For example, a multi-national company may connect a number of smaller networks together (e.g. LANs or MANs) to form a world-wide WAN. This is covered in more detail later.
• what are main benefits of networking computers and devices (rather than using a number of stand-alone computers)?

  Devices, such as printers, can be shared (thus reducing costs).Licences to run software on networks are often far cheaper than buyinglicences for an equivalent number of stand-alone computers.Users can share files and data.Access to reliable data that comes from a central source, such as afileserver.Data and files can be backed up centrally at the end of each day.Users can communicate using email and instant messaging.A network manager can oversee the network and, for example, apply accessrights to certain files, or restrict access to external networks, such as the internet.
• what are drawbacks of networking computers and devices (rather than using a number of stand-alone computers)?

  - Cabling and servers can be an expensive initial outlay.-Managing a large network can be a complex and difficult task.-A breakdown of devices, such as the file servers, can affect the whole network.-Malware and hacking can affect entire networks (particularly if a LAN is part of a much larger WAN), although firewalls do afford some protection in this respect.
• what are networked computers?

  Networked computers form an infrastructure which enables internal and external communications to take place. The infrastructure includes the following:

  hardware

  software

  services
• what does hardware contain of (as an infrustructure of networked computers) ?

  »LAN cards»routers»switches»wireless routers»cabling
• what does software contain of (as an infrustructure of networked computers) ?

  »operation and management of the network»operation of firewalls»security applications/utilities
• what do the services contain of (as an infrustructure of networked computers) ?

  »DSL»satellite communication channels»wireless protocols»IP addressing.
• what are the 2 catagories of networks?

  1) private networks

  2) public networks
• what are private networks?

  private networks are owned by a single company or organisation (they are often LANs or intranets with restricted user access, for example, passwords and user ids are required to join the network); the companies are responsible for the purchase of their own equipment and software, maintenance of the network and the hiring and training of staff.
• what are public networks?

  Public networksare owned by a communications carrier company (such asa telecoms company); many organisations will use the network and there are usually no specific password requirements to enter the network – but sub- networks may be under security management.
• what are Wireless LANs (WLANs)

  ?Wireless LANs (WLANs)are similar to LANs but there are no wires or cables. In other words, they provide wireless network communications over fairly short distances (up to 100 metres) using radio or infrared signals instead of using cables.Devices, known aswireless access points (WAPs), are connected intothe wired network at fixed locations. Because of the limited range, most commercial LANs (such as those on a college campus or at an airport) need several WAPs to permit uninterrupted wireless communications. The WAPs use eitherspread spectrumtechnology (which is a wideband radio frequency with a range from a few metres to 100 metres) or infrared (which has a very short range of about 1 to 2 metres and is easily blocked, and therefore has limited use; see Section 2.1.5 Wired and wireless networking).The WAP receives and transmits data between the WLAN and the wired network structure. End users access the WLAN through wireless LAN adapters which are built into the devices or as a plug in module.

  
• what are the 2 types of networking models?


  - client-server

  - peer-to-peer
• what is the client-server model?

  Theclient-servermodel uses separate dedicated servers and specific client workstations; client computers will be connected to the server computer(s).»Users are able to access most of the files, which are stored on dedicated servers.»The server dictates which users are able to access which files. (Note: sharing of data is the most important part of the client-server model; with peer-to- peer, connectivity is the most important aspect.)»The client-server model allows the installation of software onto a client’s computer.»The model uses central security databases which control access to the shared resources. (Note: passwords and user IDs are required to log into the network.)»Once a user is logged into the system, they will have access to only those resources (such as a printer) and files assigned to them by the network administrator, so offers greater security than peer-to-peer networks.»Client-server networks can be as large as you want them to be and they are much easier to scale up than peer-to-peer networks.»A central server looks after the storing, delivery and sending of emails.»This model offers the most stable system, for example, if someone deletesa shared resource from the server, the nightly back-up would restore thedeleted resource (this is different in peer-to-peer – see later).»Client-server networks can become bottlenecked if there are several clientrequests at the same time.

  
• In the client-server model, a file server is used and is responsible for:- central storage and management of data files, thus enabling other network users to access files

  - allowing users to share information without the need for offline devices (such as a memory stick)

  - allowing any computer to be configured as the host machine and act as the file server (note that the server could be a storage device (such as SSD or HDD) that could also serve as a remote storage device for other computers, thus allowing them to access this device as if it were a local storage device attached to their computer).
• why would a company/user choose a client-server network model for the following reasons ?

  »The company/user has a large user-base (however, it should be pointed out that this type of network model may still be used by a small group of people who are doing independent projects but need to have sharing of data and access to data outside the group).»Access to network resources needs to be properly controlled.»There is a need for good network security.»The company requires its data to be free from accidental loss (in otherwords, data needs to be backed up at a central location).
• what is a peer-to-peer model?

  On a peer-to-peer network, eachnodejoins the network to allow»the provision of services to all other network users; the services available are listed on a nominated ‘look up’ computer – when a node requests a service, the ‘look up’ computer is contacted to find out which of the other network nodes can provide the required service»other users on the network to simply access data from another node»communication with other peers connected to the network»peers to be both suppliers and consumers (unlike the client-servermodel where consumers and resources are kept entirely separate fromeach other)»peers to participate as equals on the network (again this is differentto the client-server model where a webserver and client have different responsibilities).

  
• what is 1 thing to note about peer-to-peer models?

  Thepeer-to-peermodel does not have a central server. Each of the nodes (workstations) on the network can share its files with all the other nodes, and each of the nodes will have its own data.
• why is there no requirement to authenticate users?

  Because there is no central storage, there is no requirement to authenticate users.
• in which scenarios is the peer-to-peer model used?

  This model is used in scenarios where no more than 10 nodes are required (such as a small business) where it is relatively easy for users to be in contact with each other on a regular basis. More than 10 nodes leads to performance and management issues.
• how come peer-to-peer offers little data security?

  Peer-to-peer offers little data security since there is no central security system. This means it is impossible to know who is authorised to share certain data. Users can create their own network node share point which is the only real security aspect since this gives them some kind of control. However, there are no real authentication procedures.
• Examples of peer-to-peer network model


  A user would choose the peer-to-peer network model for one or more of following reasons:»The network of users is fairly small.»There is no need for robust security.»They require workstation-based applications rather than being server-based.
• what types of client-server model are available?

  The client-server model offersthin clientsandthick clients. These can often refer to both hardware and software.
• Thin client

  A thin client is heavily dependent on having access to a server to allow constant access to files and to allow applications to run uninterrupted. A thin client can either be a device or software which needs to be connected to a powerful computer or server to allow processing to take place (the computer or server could be on the internet or could be part of a LAN/MAN/ WAN network). The thin client will not work unless it is connected at all times to the computer or server. A software example would be a web browser which has very limited functions unless it is connected to a server. Other examples include mobile phone apps which need constant access to a server to work. A hardware example is a POS terminal at a supermarket that needs constant access to a server to find prices, charge customers and to do any significant processing.
• Thick client

  A thick client can either be a device or software that can work offline or online; it is still able to do some processing whether it is connected to a server or not. A thick client can either be connected to a LAN/MAN/WAN, virtual network, the internet or a cloud computing server. A hardware example is a normal PC/laptop/tablet since it would have its own storage (HDD or SSD),RAM and operating system which means it is capable of operating effectively online or offline. An example of software is a computer game which can run independently on a user's computer, but can also connect to an online server to allow gamers to play and communicate with each other.
• What are the pros of thick-clients?

  - more robust (device can carry out processing even when not connected to server)

  - clients have more control (they can store their own programs and data/ files)
• what are the cons of thick clients?

  - less secure (relies on clients to keep their own data secure)

  - each client needs to update data and software individually

  - data integrity issues, since many clients access the same data which can lead to inconsistencies
• pros of thin clients?

  - less expensive to expand (low-powered and cheap devices can be used)

  - all devices are linked to a server (data updates and new software installation done centrally)

  - server can offer protection against hacking and malware
• cons of thin clients?

  - high reliance on the server; if the server goes down or there is a break in the communication link then the devices cannot work

  - despite cheaper hardware, the start-up costs are generally higher than for thick clients
• differences between thick and thin client software:

  Thin client points:

  - always relies on a connection to a remote server or computer for it to work - requires very few local resources (such as SSD, RAM memory or computer processing time) - relies on a good, stable and fast network connection for it to work- data is stored on a remote server or computer
• differences between thick and thin client software

  thick client software points:

  - can run some of the features of the software even when not connected to a server- relies heavily on local resources- more tolerant of a slow network connection - can store data on local resources such as HDD or SSD
• name the four network tapologies:

  »bus networks»star networks»mesh networks»hybrid networks.