Hardware Protocol Layers
LESSON 4

Outline:
Hardware Protocol Layers

General Communications Concepts and Network:

• Three types of network communication:
  • Circuit switched
  • Message switched
  • Packet switched
• LAN - Local Area Network:
  • Connect local personal computers and peripherals.
  • Usually high-bandwidth.
  • Ethernet and Token-Ring are common examples.
• MAN - Metropolitan Area Network:
  • Active switching elements
  • Short delays (distance = a few miles)
  • Local-loop
• WAN - Wide Area Network:
  • Switched packet and circuit lines
  • Commonly has a lower bandwidth than LANs
  • Significant delays in routing(distance = many miles)

Various Network Components:

• Repeaters:
  • Amplify signals.
  • Connect similar physical technologies.
  • Are used to extend the reach of LAN segments.
  • Often restricted to (at-most) four repeaters per segment.

• Bridges:
  • Function at the link-level (DLL) and provide standard link-level features.
  • Examine frames and cells.
  • Can span different physical technologies.
• Gateways and Routers:
  • Route packets
  • Connect LANs to other (physically similar or different) network types. (WANs, etc.)
  • Gateways can exist at the network-layer and higher, since they also convert protocols if needed.
• Multiplexers and Inverse MUX:
  • Combine or separate multiple data streams to or from a single stream.
  • Inverse multiplexers take a single stream and transmit it over multiple several output connections.
  • Routing or switching is not done within a MUX.
• Hubs and Switches:
  • Hubs are commonly used with LANs. They enable point-to- point connections and can also support Virtual LAN configurations.
  • Switches are primarily used for high-speed networks. The switching fabric contained within enables many connections to be switched simultaneously.
  • Switches are commonly used with broadband technologies such as ATM and Frame Relay.

Network Cabling:

• Cabling - many types of cables and connections are used to create networks. The physical characteristics of the medium dictates the type and speed of data which would be carried over it, and also the protocols which should be used.

Network Cabling:
Fiber

• Fiber - is the fastest and highest bandwidth of all current transmission mediums
• Fiber uses laser-diodes to send light-pulses at different intervals which represent binary bits.
• Electrical signals are converted to light-pulses when entering a fiber connection, then converted back to electrical signals using a photo-diode.
• Single or multimedia fiber types.

Network Cabling:
Coaxial

• Thickwire(10Base5) - used primarily in Ethernet networks requiring a 50 Ohm N-Series terminator.
  • Must be tapped using an N-Series Tap (MDI) and a coax transceiver (AUI connector).
  • Max. Length of a segment = 500 meters (1640 feet)
  • Bandwidth = 10 Mbps.
  

• Thinwire(10Base2) - popular in LANs, 10Base2 connections usually require 50 Ohm BNC terminators.
  • Must be connected using BNC connectors and joined by BNC "Tee" connectors.
  • Also referred to as "Cheapernet"
  • Max. Length of a segment = 185 meters (~200 meters)
  • Bandwidth = 10 Mbps.

Network Cabling:
Unshielded Twisted Pair

• UTP (10Base-T) - is the most popular Ethernet media because it's inexpensive, light, small, and uses the same cabling techniques as standard phone lines.
  • 10Base-T connectors are RJ-45 jacks (similar to standard phone jacks).
  • Max. Length of a segment = 100 meters (up to 150)
  • UTP is provided in 5 different classes (or levels) which determine the type and speed of data the cabling can carry.

Network Cabling:
Wireless

• Wireless networks are becoming increasingly popular in today's industry. Different wireless technologies exist to support most networks.
  • Satellite - high-orbit and geo-stationary satellites provide WANs with connectivity.
    • An average 250ms delay is introduced by satellite transmission.
  • Radio - towers provide land to land wireless links that are commonly used in cellular networks.
  • Infrared wireless connections are sometimes used for campus- wide networks, transmitting and receiving infrared rays from building-tops.

Network Interface Cards:

• Network Interface Cards - contain the network protocols that allow a computer to transmit and receive information over a physical transmission medium.
  • NICs conform to the hardware (lower-layer) protocols which dictate how it should:
  • Transmit raw bits onto a wire
  • How it should access the medium (collision detection, token)
  • Encoding techniques and frame delimiters

Topologies:

• Star - direct connection from a single center node to each surrounding node (complete bandwidth).
• Ring - can be point-to-point (store and forward) or nodes can simply be attached to a ring cable (complete bandwidth).
• Bus - all nodes share the same transmission medium and the same bandwidth. Collisions during transmission can occur.
• Mesh - all nodes have a direct link to every other node (expensive, wasteful, complete bandwidth).

Conceptual Protocol Layering:

• When protocols are combined to provide a communications suite, they are commonly stacked at different conceptual layers.
• At different layers, each protocol performs an unique task or function required to make communication possible.
• The interoperability and functional independence of communication protocols allows them to be interchanged without a reduction in functionality.
  

• Protocol Data Units (PDU) - are the standard unit of information that is passed between protocol layers.
• Encapsulation - The technique used by layered protocols in which a layer adds header information to the protocol data unit (PDU) from the layer above.
• For all conceptual layers:
  • Layer N treats a PDU from layer N+1 strictly as data (including headers and all).
  • Layer N treats a PDU from layer N-1 as separate parts:
    • Header part - contains information it needs to process.
    • Data part - to be sent to the destination (layer N+1) dictated by the header information.
• Control information and PDUs are passed between layers via service access points (SAP).
• Some protocol layers may add trailers as well as headers. This usually happens at the link layer, where data may need to be framed.

Open Systems Interconnection
(OSI) model:

Open Systems Interconnection:
Layers

• Lower-Layers:
  • The lower conceptual layers directly interact with the hardware and physical transmission medium.
  • Commonly, the lowest two layers are directly involved (depending upon the protocol stack).
    • Layer 2 - Link Layer
    • Layer 1 - Physical Layer

Open Systems Interconnection:
Physical Layer

• The physical layer is not the transmission medium itself, but the conceptual layer which directly interacts with it.
• It may contain several sublayers and, as a whole, is responsible for:
  • Transmission of raw bits as different voltages.
  • Pulse duration of the bit.
  • Type of transmission (half-duplex or full-duplex)
  • Specifications regarding the network connector.

Open Systems Interconnection:
Data Link Layer

• The link layer protocol acts on a point-to-point basis inside a subnet and most often is made up of sublayers, including a Medium Access Sublayer (MAC).
• The link layer (LLC, HDLC) usually handles:
  • Error detection and correction.
  • Construction of frames (assembly and reassembly).
  • Discarding of duplicate frames.
  • Flow control.

Open Systems Interconnection:
MAC Sublayer

• The MAC sublayer usually determines the access method to use for the underlying transmission medium.
• The MAC sublayer includes standards such as:
  • IEEE 802.3 - Ethernet CSMA/CD (Carrier Sense Multiple Access / Collision Detection)
  • IEEE 802.4 - Token-bus specification.
  • IEEE 802.5 - Token-ring specification.

Open Systems Interconnection:
Network Layer

• The network layer deals with an end-to-end connection inside a subnet, not source and destination.
• It concerns itself with:
  • Routing packets.
  • Controlling congestion in the subnet (global issue).
  • Performing network accounting functions.
  • Providing internetworking.
  • Signaling

Old Lower-Layers Model:
OSI

New Lower-Layers Model:
ATM

• Layer 1 (Physical):
  • SONET (Synchronous Optical Network)
  • SDH(Synchronous Digital Hierarchy)
• Layer 1.5 (MAC):
  • ATM Layer
  • Handles medium-dependent functions.
  • Multiplex and demultiplex cells.
  • Perform virtual circuit and path translations.
• Layer 2 (Data Link):
  • ATM Adaptation Layer
  • New lowest end-to-end layer
  • Performs standard link layer functions.
  • Contains sublayers:
    • Convergence Sublayer (CS)
    • Segmentation and Reassembly Sublayer (SAR)

Ethernet (IEEE 802.3)

• Xerox/PARC Ethernet specification influenced the appearance of the 802.3 standard. The IEEE standard is commonly referred to as Ethernet when in reality it’s only based on it.
• CSMA/CD (Carrier Sense Multiple Access / Collision Detection) - is the access method used by IEEE standard in which it must detect transmission collisions and retransmit after a random back-off time.
• Delays can be caused by collisions (not sensitive to real-time traffic).

Ethernet (IEEE 802.3):
Ethernet Addresses

• Ethernet uses 48-bit NIC hardware addresses which are uniquely defined and maintained by IEEE.
• Physical addresses are associated with the Ethernet interface hardware; moving the hardware interface to a different machine or changing the card in the current machine, changes the machine's physical address.
• 48-Bit Ethernet addresses also allow for (beyond single node addressing):
  • Multicast addresses
  • Broadcast addresses

Ethernet (IEEE 802.3):
Ethernet frames

• Link-level Ethernet frames are self-identifying because of the 16-bit FRAME-TYPE field and their unique hardware addresses.
• Chief advantages:
  • Allow multiple protocols to be used together on the same machine.
  • Allow intermix of multiple protocols on the same physical network without interference.

Token-based Standards:

• Token Bus (IEEE 802.4)
  • Bus topology over which tokens are used as the medium access method.
  • Not very popular in today's LANs.
• Token Ring (IEEE 802.5)
  • The token-ring specification was created out of a need for bounded round-trip times.
  • Switched Token-ring topologies and devices are becoming increasingly popular.

IEEE Ethernet-related standards:

• isoEthernet (IEEE 802.9a)
  • Provides a way to efficiently deliver voice, video, and data- streams to the desktop over 10Base-T (UTP) two-pair Class-3 cabling.
  • Uses TDM (time division multiplexing) over 56K/64Kbps point-to-point channels through a switched network.
  • Bandwidth = one 10Mbps channel for 10Base-T packet data and one switched 6.144Mbps (96 - 64Kbps ISDN B-channels) for isochronous traffic.
• Switched Ethernet
  • Provides complete access to all 10Mbps of bandwidth over a point-to-point connection.
  • Limited by the number of back-plate ports and the switching fabric.
• 100Mbps Fast Ethernet (IEEE 802.3u)
  • 100Base-X is the true heir to the 802.3 Ethernet standard by providing 100Mbps of network bandwidth.
  • Fast Ethernet utilizes a media independent interface (MII) which enables it to support various cabling types on the same network.
• 100 VG-Any LAN
  • Not officially an Ethernet-related product, VG-Any LAN provides 100Mbps of bandwidth for local area networks.
  • In direct competition with Fast Ethernet, VG-Any LAN has failed to capture any substantial percentage of the high- speed networking marketplace.

FDDI:

• Fiber Distributed Digital Interface (FDDI) - is commonly used as a backbone for campus-wide networks and some MANs.
  • It provides 100Mbps of bandwidth for up to 1000 connections.
  • FDDI uses dual-rings which provides a recovery mechanism in case a node crashes ("self healing").
  • An FDDI ring can span up to 100 kilometers.
  • FDDI technology uses a token as it's medium access control.

FDDI-II:

• FDDI only supports data traffic. To carry voice connections, FDDI-II was created.
• A cycle in FDDI-II spans (125microsec) which enables it to support real-time voice connections multiplexed together with data.
• Since both data and voice are supported, FDDI-II uses cycle groups to delimit where data and voice information begins and ends.

N-ISDN:

• Integrated Services Digital Network (ISDN) - supports a wide range of voice and non-voice applications. ISDN can handle a range of services over a common digital network.
• Narrowband ISDN - is simply referred to as ISDN at primary rates or lower.
• N-ISDN is available at various service rates. The two most popular are primary rate and basic rate.
• Each service rate contains two types of channels:
  • Digital data channels (B) - which provide 64 Kbps per channel.
  • Signaling channels (D) - which provide out-of-band control information at rates of 16 or 64 Kbps.
• Basic Rate provides:
  • 2 B-channels at 64 Kbps.
  • 1 D-channel at 64 Kbps.
  • Maximum available throughput = 192 Kbps.
• Primary Rate (U.S.) provides:
  • 23 B-channels at 64Kbps.
  • 1 D-channel at 64Kbps.
  • Maximum available throughput = 1.544 Mbps. (T1)
• Primary Rate (E.C.) provides:
  • 30 B-channels at 64Kbps.
  • 1 D-channel at 64Kbps.
  • Maximum available throughput = 2.048 Mbps. (E1)

Last Modification: (Sunday, August 25, 1996)

All work was written, produced, and is copyrighted by Daniel Z. Tabor Jr.
Page created by Daniel Z. Tabor Jr. Copyright ©1996 Illusion Industries Inc.