Data networks are telecommunications networks installed and operated exclusively for information exchange between data communication devices (such as computers). Data network types include premises distribution network (PDN), local area networks (LAN’s), metropolitan area networks (MAN’s), and wide area networks (WAN’s). These are hierarchical with the LAN being the base and the WAN being the umbrella architecture.
PANs are short-range data communications systems that are primarily used to interconnect peripheral equipment with a local computer or computing system. LANs are designed to reliably transfer large amounts of data quickly and error-free over a very small area such as an office. MAN’s facilitate LAN-to-LAN information exchange in a local exchange area. WAN’s allow for information exchange between LAN’s in different exchanges normally across LATA boundaries. For example a LAN in Chicago sharing information with a LAN Seattle would do so across a WAN.
A data network is composed of several key parts such as data terminals (e.g., personal computers), network adapters, access wiring, and data distribution nodes (e.g., routers, brouters, and switches). In some data networks, network management/control systems are used to configure, monitor, coordinate, and control the network elements.
Data Terminals
Data terminals are data input and output devices that are used to communicate with a remotely located computer or other data communication device. Data terminals frequently consist of a keyboard, video display monitor, and communication circuitry that can connect the data terminal with the remotely located computer.
The term “data terminal” is often used to describe multiple types of devices including personal computers (PCs), dedicated “dumb” terminals, scientific workstations, and other types of computers that can communicate with other computers or a host computer.
Data terminal equipment (DTE) are devices that capture and serialize information for communication to other communication devices. Data communication equipment (DCE) circuits are assemblies that convert data information into a format that can be transferred through a communication network.
Figure 1 shows data terminals that are connected through a modem to interconnect the data terminals with a remote computer. In this diagram, the data terminals are the DTE and the modems are the DCE.
Network Interface Card (NIC)
A NIC is a device that adapts the data communication network protocol to a data bus or data interface in a computer. The NIC is installed between a computer network (such as the Ethernet) and a computer data bus (such as a PCI socket). The NIC is usually a PC expansion board connector and operating system. Software in the computer is installed and setup to recognize the NIC card.
LAN Wiring
There are typically three types of wiring used for LAN’s: twisted pair, coax, and fiber. Of these, twisted pair is dominant for several reasons: ease of installation, availability, cost, and speed as a function of relative cost.
Twisted pair comes in a variety of “categories” and is either shielded twisted pair (STP) or unshielded twisted pair (UTP). UTP is the less expensive and the most widely used. STP has an outer copper or foil conductor located just beneath the out sheath of the wire. In areas where there is a significant incidence of electromagnetic interference (EMI), such as around factory floor machinery or hospital radiological/MRI equipment, STP is used.
Twisted pair wire is classified by categories that relate to the data transmission speed at which the wire is capable of passing data. For each category there are manufacturing specifications such as wire quality, insulation characteristics, and number of twists per inch. Generally, the higher the number of twists, the higher the data transmission rate can be.
Routinely LAN cable is four-pair (eight conductors) even though most data communication systems (such as Ethernet) only require 2 pairs (transmit and receive pairs). It is installed with all conductors terminated on each end into patch fields, hub equipment, or office wall plates (jack fields). From the office wall jack the typical PC or peripheral device is connected to the LAN via a wall cord that is also four-pair terminated in RJ-45 modular connectors. Most offices are wired for multiple network connections and in many cases the voice and data wiring is installed together and to the same cable specification (e.g., category 3 and above).
Network Distribution and Routing
Network distribution and routing equipment provides communication paths between the end-user and the services they desire to use (e.g., Internet). There are three basic methods used to distribute in data networks: broadcast (distribution hubs), dedicated paths (switching nodes), and packet-switching (routers).
Hubs broadcast information to all the communication devices that are connected to it. Switches create a physical or logical connection between data communication devices. Routers are intelligent switches that can dynamically route (switch to other routers) packets of data toward their ultimate destination.
Network Access Control
Network access control is a process of coordinating access of data communication devices to a shared communications media (transmission medium). Network access control is a combination of media access control (MAC) and service authorization.
There are two key ways data communication devices can access communication systems: non-contention based and contention based. Non-contention based regularly poll or schedule data transmission access attempts. An example of a non-contention based data communication system is token ring. In the token ring system, only the data communication device that has the token is allowed to transmit. This ensures that other data devices will not interfere with the data transmission. Contention based access control systems allow data communication devices to randomly access the system through the sensing and coordination of busy status and detected collisions. Carrier sense multiple access (CSMA) with collision detection (CSMA/CD) or collision avoidance (CSMA/CA) listen to the data activity first to determine if the systems is not busy (carrier sense) before they begin a transmit request. After the device transmits its required, it waits to hear if the system has acknowledged its required (usually an echo of its original signal). If the CSMA/CD device does not hear an acknowledgement, it will wait a random amount of time before transmitting another data transmission service request.
The CSMA/CA system differs from the CSMA/CD system by the assignment of different access wait periods to different priority groups of devices. This allows high priority devices (such as a system management data terminal) to attempt access before a lower priority device (e.g., web browsing terminal).
Figure 2 shows the key ways networks can control data transmission access: non-contention based and contention based. This diagram shows that non-contention based regularly poll or schedule data transmission access attempts before computers can begin to transmit data. This diagram shows that a token is passed between each computer in the network and computers can only transmit when they have the token.
Because there is no potential for collisions, computers do not need to confirm the data was successfully transmitted through the network. This diagram also shows contention based access control systems allow data communication devices to randomly access the system through the sensing and coordination of busy status and detected collisions. These devices first listen to see if the system is not busy and then randomly transmit their data. Computers in the contention-based systems must confirm that data was successfully transmitted through the network, because there is the potential for collisions.
No comments:
Post a Comment