Networking Fundamentals

Building Ethernet LAN's

Ethernet was developed in the 1970's at Xerox under Robert Metcalfe and commercially introduced in 1980 and standardised in 1983 as IEEE 802.3. These standards defined rules for such things as "Frame Format" as well as "how computers communicate" over a single wire shared by all devices on the network

Ethernet operates at Layer 1 and Layer 2. Both of these layers are implemented on the NIC (network Interface Card. The Ethernet PHYSICAL layer defines such things as 'connection speed' and 'type' of cabling

Ethernet Components:

CSMA/CD - Carrier Sense Multiple Access with Collision Detection. Only used when devices are in half duplex mode (Bridges/Hubs/Repeaters). Collisions occur when on half-duplex

How do devices know if its their turn to send data (think of a group of people in a room - listen, then sends data if no one is speaking - CSMA/CD regulates the conversation - TX OR RX but not at the same time. Half duplex = like walkie-talkie and full duplex like telephone, TX and RX simultaneously.

Ethernet II Frame Fields

The Ethernet packet starts with a seven octet preamble and one-octet Start Frame Deliminator (SFD). The preamble consists of a 56 bit (7-byte) pattern of alternating 1 and 0 bits, allowing devices on the network to easily synchronise their receiver clocks, providing bit-level synchronisation.

It is then followed by the SFD to provide byte-level synchronisation and to mark a new incoming frame

The SFD is the eight-bit (one-byte) value that marks the end of the preamble

The TYPE field identifies which LAYER 3 PROTOCOL is being transported in the Ethernet Frame eg IP, Appletalk, IPX etc.

Early Ethernet

Ethernet has strict rules on the lengths of physical cable segments as well as how many devices on the cable. To extend this length, REPEATERS (regenerate the signal) were introduced.

• Repeaters: Are PHYSICAL layer devices and were used to connect 2 or more physical cable segments, making it act like ONE long cable

  • Lengths:

    • Fibre Optics >10Km

    • Thick Ethernet (10Base5) =500m

    • Thin Ethernet (10Base2) = 185m

    • Twisted Pair (10Base-T) = 100m – Cat 3

    • Fast Ethernet(100Base-T)=100m – Cat 5/5e

    • Gigabit Ethernet (1000B-T) =100m – Cat 6/6a/7

• 5-4-3 Rule (Still 1 Collision Domain)

• Hubs: Also PHYSICAL layer components but differ from Repeaters as they use a "STAR" configuration as opposed to the "SHARED BUS" cable setup that Repeaters use. The shared bus is is still in use BUT its now contained within the Hub. The other advantage that Hubs have is that if there is a cable break, only one user is affected as opposed to the repeater where everyone is affected. This is still a half duplex setup and has 1 collision domain (therefore uses CSMA/CD)

  As more and more people connected to the collision domain, performance issues arose.To get around these issues, was to break up the large collision domain into separate collision domains

• Bridges: These devices connect 2 or more physical cable segments to create a larger network. Unlike a repeater or hub, each side of the bridge becomes a separate collision domain or network segment. This was called 'network segmentation' where collisions are isolated withi the respective segments

Bridges are Layer 2 devices and use MAC addresses to forward traffic. In effect Bridges does the following: 1) Forwards Frames using the DESTINATION MAC address to make decisions 2) Learns MAC addresses 3) Controls Traffic An issue arises whereby multiple bridges are connected to increase the size of the network as well as to provide REDUNDANT links. Should a link fail, an alternative route can be taken to the destination. The problem is that these redundant links can cause LOOPS in the network whereby bridges continually update their MAC address table with the same MAC address being received on DIFFERENT ports. To mitigate these loops (from broadcast and multicast messages) bridges use SPANNING TREE PROTOCOL (STP) and puts ports into "BLOCKING" mode and others into "FORWARDING" mode

Wireless LAN Architectures: There are 3 common WLAN architectures:

• Autonomous: This is known as FAT AP (or FAT Access Point). This is where ALL functionality is coded into every AP. Home networks fall into this category

• Centralised or Switched: This is known as Switched WLAN System. Here the AP's become 'thin' or 'lightweight' and there is central switch or AP that manages and controls the AP's. These AP's do not store configuration data or credentials etc, This is all done by the centralised management switch which provides functionality such as security, VLAN's, user roaming, QoS etc. Most enterprises use the "Centralised or Switched" WLAN model due to its scalability.

• Distributed: This is known as the "Wireless Meshed Network", whereby using cables would be difficult (such as open air functions etc) Data is repeated over wireless to neighbour nodes. Like a network of Routers, the devices make forwarding decisions to get the data through the network as efficiently as possible.

• 802 .11 Group of Standards:

  • 1997 IEEE 802.11 standard was released for WLAN's

    • 802.11 legacy (1&2 Mbps @ 2,4GHz using DSSS and FHSS

  • 1999 Wi-Fi brand was adopted

    • 802.11b (5-11Mbps @ 2,4GHz using DSSS)

    • 802.11a (54 Mbps @ 5GHZ using OFDM)

  • 2003

    • 802.11g (54 Mbps @2,4 GHZ using DSSS and OFDM)

  • 2009 - Wi-Fi 4

    • 802.11n (600 Mbps @2,4 & 5 GHz using OFDM)

  • 2014 - Wi-Fi 5

    • 802.11ac (3.46 Gbps @5GHz using OFDM)

  • 2016

    • 802.11ah (347 Mbps@908-928 MHz)

  • 2020 - Wi-Fi 6

    • 802.11ax (9,6Gbps @ OFDMA)