|
A repeater copies a signal received from one port out through its other port. This effectively lengthens a cable.
Hubs are multiport repeaters that replicate a signal received from one port out through all other ports. There is no filtering.
A hub is the central piece of equipment in a physical star topology. It is used in 10BaseT, 100BaseTX, and 1000BaseT networks.
A passive hub will repeat a signal but won't regenerate it. An active hub will regenerate and amplify the signal.
Each computer connects to the hub with its own cable.
|
|
A bridge connects two LAN segments but only passes through network packets that are destined for the other segment.
A bridge can reduce traffic problems by filtering out the network packets between the two connected segments.
A transparent bridge is used in Ethernet networks. This type of bridge lists the Media Access Control (MAC) address of all network devices by way of a tag that states which segment the device is located on.
A source route bridge is used in token ring networks. This type of bridge expects all network devices to store a copy of the table of devices in their memory.
|
|
A switch is essentially a multiport bridge. Each port on a switch can be connected to a single device or an entire network segment.
Data received from one port on a switch is forwarded out only on the port that has the device for which that data is destined. (Compare this to a hub that forwards data to all ports.)
Both bridges and switches forward data based on the destination MAC address of the data.
|
|
A router connects two or more networks and forwards only the data that is destined for a particular network.
A router will forward data based on the logical network segment and node address of the data packet.
Routers can connect dissimilar network architectures.
A Brouter provides the same function as a router but also bridges nonroutable protocols. A Brouter is a combination of a router and a bridge.
|
|
A gateway connects networks that use entirely different protocols.
A gateway acts as a translator between the two protocol stacks. For example, a gateway can translate between IPX/SPX on one network and TCP/IP on the other.
|
|
A CSU/DSU is used in WANs and converts digital data frames for the WAN to the protocols used on the LAN.
CSU/DSUs are often used with leased lines.
|
|
NICs, ISDN (Integrated Services Digital Network) adapters, and SAN adapters are expansion boards that are placed within a computer to enable it to communicate with external equipment.
NICs provide the ability for the computer to transmit data in packet form onto a network.
ISDN adapters allow a computer to communicate across the ISDN, which is a digital network used by the telephone company.
SAN adapters allow a computer to communicate with a storage area network (or system area network) to access or exchange stored data.
|
|
A WAP is similar to a hub for an Ethernet network except that it transmits in radio frequences across air, rather than as digital signals across copper cables.
WAPs are used in wireless networks in which more than two devices connect to the network via wireless transmissions.
|
|
A modem is a modulator/demodulator that modulates digital signals into analog signals, transmits them across the plain old telephone system (POTS), and then demodulates analog signals received from the telephone system.
External modems are connected to the computer via an RS-232 serial connector.
|
|
|
|
[Linkafbeelding 0300.gif]
Tel de zwarte stippen van de achtergrond afbeelding!
|
|
The term network topology describes how a network is connected. A physical topology defines the physical connections. A logical topology defines the shape of how data travels throughout the network.
Logical and physical topologies are not necessarily the same. For example, a token ring network uses a physical star topology, although data travels around the star in a logical topology shaped like a ring.
A topology can be applied to a local area network (LAN), which is a typical network inside a building or office; a metropolitan area network (MAN), which is a network consisting of several LANs within a metropolitan area such as a city; or a wide area network (WAN), which is a network spanning a long distance and consisting of two or more LANs.
|
|
Description Each device is connected to a central hub.
Uses 10BaseT, 100BaseT.
Advantages Easy to troubleshoot, install, and add to. Cabling is cheap.
Disadvantages Centralized point of failure (hub).
---
Star topologies are physical only. They consist of many cables connected to a central hub.
A failure of a single cable, workstation, or hub port will not affect other devices.
A failure of the hub will affect all connected devices.
Star topology has a central hub with separate cables to each device.
|
|
Description Each device is connected to a common backbone.
Uses 10Base2, 10Base5, cable TV.
Advantages Easy to install, very cheap, uses the least amount of cable.
Disadvantages Difficult to troubleshoot, no fault tolerance.
---
The bus topology consists of a single cable trunk to which all nodes are connected.
The length of the trunk is used to measure the total length of the cable; the short drop cables connecting the nodes to the trunk are not added into the total length of the cable.
Each end of the cable trunk must be terminated.
10Base5 and 10Base2 networks use both a physical and logical bus topology. A 10BaseT network uses a physical star topology with a logical bus. In a logical bus topology, a node sends data to all other nodes simultaneously.
The advantage of a physical bus topology is that it is cheap and simple to install.
The disadvantage of a physical bus is that a failure affects all other nodes, and access traffic can occur because data is passed to all nodes.
A physical bus topology has a main cable with each node attached to it by short drop cables.
|
|
Description Each device has a point-to-point connection to every other device.
Uses Used in some WAN implementations.
Advantages Very fault tolerant.
Disadvantages Expensive and difficult to set up.
---
The mesh topology consists of each node connected to all other nodes.
The main advantage of a mesh topology is the redundancy it provides, which preserves connectivity even in the event of a failure between links.
The disadvantage of a mesh topology is that it is very expensive to set up.
Each node is connected to all other nodes in a mesh topology.
|
|
Description Each device is connected to two others to create a closed loop.
Uses Token-ring, FDDI.
Advantages High speed, easy to troubleshoot.
Disadvantages Expensive hardware.
---
The physical ring topology is a group of nodes that are joined in daisychain fashion until the last is connected to the first, creating a ring of communication.
A logical ring topology transmits data in a ring from node to node, even if the physical topology is a star topology, as in a token ring network.
The advantage of a ring topology is that it provides equal access to the network.
A disadvantage of a ring topology is that a failure in the ring will affect all other nodes in the ring.
A physical ring topology is created by connecting each node to the next until the last is connected to the first.
|
|
Description Each device accesses other devices via radio transmission.
Uses 802.11b.
Advantages Mobile solution
Disadvantages Difficult to troubleshoot, slow speed.
---
Wireless topology (also called cellular is one in which a wireless access point (WAP) creates a cell within which any wireless device can connect and WAPs can be interconnected.
An ad-hoc wireless network can be created by connecting two wireless devices directly to each other without the use of a WAP.
The advantage of wireless networks is that devices can easily move about and retain connectivity.
The disadvantage of wireless networks is that they are subject to interference. In addition, wireless networks will degrade in speed as a device moves farther away from the WAP (or other device, in the case of an ad-hoc network).
Wireless topology is typified by cells of connectivity centralized at the WAP.
|
|
|
The Institute of Electrical and Electronics Engineers (IEEE) 802 specifications are for frame types. A frame type is the format of the packet that is used to communicate across the network.
The 802.2 specification provides for the Logical Link Control (LLC) header information to identify upper layer protocols.
Specification 802.3 or Ethernet
Speed 5 Mbps, 10 Mbps, 100 Mbps
Access Method Carrier Sense Multiple Access with Collision Detection (CSMA/CD)
Topology Physical bus, physical star, logical bus
Media Copper unshielded twisted pair (UTP), copper coaxial
---
Specification 802.5 or token ring
Speed 4 Mbps, 16 Mbps
Access Method Token passing
Topology Physical star, physical ring, logical ring
Media Copper UTP, copper shielded twisted pair (STP)
---
Specification 802.11(b) or Wireless
Speed 11 Mbps
Access Method Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA)
Topology Wireless
Media Air
---
Specification 802.12 or 100VGAnyLan
Speed 100 Mbps
Access Method Demand priority
Topology Physical Star, logical bus
Media Copper UTP
---
Specification FDDI or Fiber Distributed Data Interface
Speed 100 Mbps
Access Method Token passing
Topology Physical ring, logical ring
Media Fiber optic
|
|
The Ethernet IEEE 802.3 specification can run across fiber optic, coaxial, and UTP copper cabling.
To use coaxial cabling in a physical bus topology (which can be either 10Base2 or 10Base5), you must follow the 5-4-3 rule, which states that there can be five (5) cable segments connected with four (4) repeaters, but only three (3) of the segments can be populated with nodes.
10BaseT
10BaseT uses copper UTP cabling of Cat3 or better grade, connected to a central hub, to create a physical star topology.
The UTP cables in a 10BaseT network use a registered jack 45 (RJ-45) type of connector.
10BaseT runs at 10Mbps, and cables can be a maximum of 100 meters in length. This lenght includes the spans of the patch cable and drop cable.
100BaseTX
On the exterior, 100BaseTX looks the same as 10BaseT because 100BaseTX uses copper UTP cabling of Cat5 or better grade connected to a central hub. Cat5 has two pairs of copper wires within the cable.
All cables use an RJ-45 connector and have a maximum cable length of 100 meters.
100BaseTX provides speds of up to 100 Mbps.
10Base2
10Base2 uses a physical bus topology in which all computers connect to a maion cable, which is RG-58 coaxial cable, also called ThinNet.
The maximum length of the ThinNet coax cable is 185 meters.
A 10Base2 network allows only 30 computers per segment and provides a speed of 10 Mbps.
10Base5
10Base5 uses a physical bus topology in which all the computers connect to a main cable, and it requires RG-8 or RG-11 coaxial cable, which is known as ThickNet.
The maximum length of a 10Base5 segment is 500 meters, and it offers 10 Mbps.
100BaseFX
100BaseFX provides Ethernet over fiber optic cable at a 100-Mbps speed.
The maximum length of the fiber optic segment is 6,562 feet, or 2000 meters.
Gigabit Ethernet
Gigabit Ethernet uses Cat5 or better grade cabling in a physical star topology and requires RJ-45 connectors.
Gigabit Ethernet is also called 1000BaseT, and it offers 1 Gbps of speed.
|
|
Connector RJ-11
Cabling Telephone unshielded twisted pair (UTP)
Usage Used for telephones
---
Connector RJ-45
Cabling UTP, Cat3, or better grade
Usage Used in Ethernet 10BaseT, 100BaseT, 1000BaseT, and token ring over UTP networks
---
Connector AUI
Cabling ThickNet RG-8 or RG-11 coaxial cabling
Usage Used in 10Base5 networks
---
Connector BNC
Cabling ThinNet RG-58 coax
Usage Used in 10Base2 networks
---
Connector ST and SC
Cabling Fiber optics
Usage Used in 100BaseFX networks
|
|
Each type of media, whether it is UTP, fiber, or coax, varies from the others from the standpoint of cost and installation.
You can't mix cables. In a 10Base5 network, you can't use ThinNet coax for all or even part of the backbone cabling (although you might use ThinNet for the drop cables with the appropriate transceivers). In a network that requires a certain grade of UTP (for example, 100BaseT requires Cat5 or better grade), you must use only that grade of UTP throughout the network.
The cabling must be correct for the network interface card (NIC).
The cabling must be correct for the hub and network type. For example, you can't use telephone wire with a 100BaseT network.
The connectors must be appropiate for the cable, NIC, and network type.
|
|
0
