Documentation and Theory of the
Network Construction Kit 
(Markus Kobel, Florence Lüthy, Marc Mooser)

Table Of Contents

     1.0  Introduction
          1.1  Cables and Networks
               1.1.1  Cables
          1.2  Devices
              1.2.1  Repeater
              1.2.2  Bridge
              1.2.3  Router
          1.3  Ethernet Topology
          1.4  CSMA/CD
          1.5  Network (Interface) Cards
     2.0  Topologies and Internet Protocol
          2.1  IP Address
     Multiple choice questions (solutions)
     Appendix  Books and References in the Internet
 

1.0 Introduction 

In this article we discuss Ethernet (IEEE 802.3 standard), which is the most popular way to build a LAN (Local Area Network) today. There are other LAN types like Token Ring, Fast Ethernet (IEEE 802.3u), FDDI (Fiber Data Distributed Interconnect), LocalTalk or ATM (Asynchronous Transfer Mode).
 

1.1 Cables and Networks 

Ethernet Network	Topology	Cables, Plugs and other components	Speed
10Base-T, 100Base-T	star, peer-to-peer	Unshielded Twisted Pair, RJ45 (Registered Jack)	10Mbps or 100Mbps
10Base-2 (internet-link)	bus	thin coaxial cable, BNC (Bayonet-Neill-Concelmann) Connectors, T-Piece, Terminator	10Mbps maximum
10Base-5	bus	Yellow Cable = thick coax = Thickwire, Transceiver, Terminator	1, 2, 5 or 10Mbps
10Base-F		fiber optic	10Mbps

The notation 10Base-5 for example means that it operates at 10 Mbps, uses baseband signaling and can support segments of up to 500 meters.
The most interesting network type is the 10/100Base-T network because it`s the most used one. The "T" in 10BASE-T stands for "twisted" in reference to the twisted-pair cable.

1.1.1 Cables 

There are also some combinations of these types available. For example:
    - UTP (unshielded twisted pair)
    - STP (shielded twisted pair)
 

-->show/hide typical Ethernet (RJ45) Plug (www.cyberwalker.net/connectors/ images/network-plug.jpg)

--> RJ45 connectors look (almost!) like phone connectors (help.exeter.edu/ITImages/)

--> a PCMCIA card for laptops with plugs (connectors) for both, coaxial and RJ45 cables (http://www.lll.uiuc.edu/g8/laptop/images/lancard_cable.jpg)
 
 

1.2 Devices

To give an overview, the following list contains the most important devices which can get connected:
           - PCs (Servers, Hosts, Workstations)
           - Network (Interface) Card (NICs)
           - Hubs
           - Switches
           - Bridges
           - Switching Hubs
           - Routers
           - Repeaters
Each device has its own place in the network topology and therefore its own task in the network.

1.2.1 Repeater

A repeater works on layer 1 of the ISO/OSI reference model. It regenerates the signal before sending it to the port. There are also often used multi-port repeaters, which are not very intelligent - it passes forward all received packets to all ports.
For example in 10Base2-Networks no repeaters were used, but the hosts were connected in a row and it was difficult to detect cable breaks. 10BaseT Networks which are using repeaters allow us to find cable breaks immediately.

1.2.2 Bridge

A bridge works on layer 2 of the ISO/OSI reference model to connect two network segments together. There are also commonly used multi-port Bridges, which are often called switches.
A bridge is more intelligent than a repeater. That means that the bridge knows the connected devices and sends the received packets only to the right port.

1.2.3 Router

A Router works on layer 3 of the ISO/OSI reference model.
In difference to Repeaters and Switches, Routers do actively filter incoming data before passing them forward.

1.3 Ethernet Topology 

Ethernet networks are the most commen used networks for LANs (local area networks). An Ethernet is normally built with copper cables. The topologies of 10Base-T Ethernet networks look very much like trees. The next picture will show some example configurations.

--> show/hide Topology Examples

Just to give you an idea of what Ethernet Networks are about (http://www.lantronix.com/learning/tutorials/etntba.html)
The simplest way to build a network is just to join two hosts (or PC's). In this case a crossed cable is used.

1.4 CSMA/CD

The CSMA/CD (carrier sense multiple access with collision detection) mechanism is often used in todays Ethernets. CSMA/CD is a nondeterministic access method. An example of the sending process:
The station which wants to send, first checks the momentanous network activities (Carrier Sense). If it detects activities from other stations, it waits. If there is no activity on the network, the station begins to send after a defined waiting time. Afterwards the station checks the medium for a predefined time. If it sees, that its transmission has had a collision with another transmission of another sender, it stops the own transmission and sends a so called 'Jam-Signal' to inform all other stations about the collision. After a random period of time the stations begins to send again in the same way as before.
A random period of time is important. Otherwise two stations could send permanently at the same time and collisions would be predetermined.
 

1.5 Network (Interface) Cards

--> show/hide example of a Network Interface Card (Nic)

             example NIC with BNC T Adaptor
                  (http://www.usbcable.com/networking/29-7785.htm">http://www.usbcable.com/networking/29-7785.htm)

 

2.0  Topologies and Internet Protocol 

When you are joining two different nets with the help of a router, it's important that both nets have different net-IDs.

2.1 IP-Address

You may also think that the address is devided in only two parts, which will make it sometimes easier to understand; especially when the address is in decimal form. Then the class bits would be a part of the "beginning bits" (bits before the host-ID), and these bits can be always written as x, x.x, x.x.x or x.x.x.x in decimal form.

IP-Address:
 

class bits   (1 up to 4 bits)

net-ID    (7, 14 or 21 bits)

host-ID    (24, 16, or 8 bits)

"beginning bits" (x, x.x, x.x.x or x.x.x.x)

host-ID (x.x.x, x.x, x or -)

Examples:
 

decimal	binary   (class-bits,net-ID-bits and host-ID-bits)	class	number of net-ID bits / host-ID bits
44.123.110.224	00101100  01111011  01101110  11100000	A	7 / 24
129.6.48.100	10000001  00000110  00110000  01100100	B	14 / 16
128.240.1.109	10000000  11110000  00000001  01101101	B	14 / 16

The address-class determines the possible valid beginning-part of the address:
 

address class	class bits	number of net-ID bits / host-ID bits	maximal number of nets (or net-numbers)	maximal number of host for one net (or net-number)	first valid value for the beginning-part of the IP-address (class bits and net-ID bits)	last valid value for the beginning-part of the IP-address (class bits and net-ID bits)
A	0	7 / 24	126          (27-2)	16.777.214  (224-2)	1	126
B	10	14 / 16	16.382     (214-2)	65.534         (214-2)	128.1      (128=10000000)	191.254  (191 = 10111111)
C	110	21 / 8	2.097.150  (221-2)	254               (28-2)	192.0.1	223.225.254
D	1110	-			224.0.0.0	239.255.255.254
E	1111	-			240.0.0.0	255.255.255.254

Appendix  Books and References into the Internet  

        Books

        Internet-Links

             http://www.lantronix.com/learning/tutorials/

             http://www.ethermanage.com/ethernet/ethernet.html