. . __ __._____._____:_____._____._____._____:_____._____._____ : Y : _ : _ Y_ _ : _ : . : ___Y __ : . : ____: --------| _ | _ | | _/ ___j _ | | l___ | \_ | | | \_ |-------- _\\//_ | : : : : : | L : | :__: : | | / : | : | : _\\//_ (_o,O) `--'--'--'--'--'--^-----'--'--' `--'-----^-----'-----'-----' (O.o_) " " Some info on Local Area Networks and Ethernet use in the TXEE " " Psyclone of MED in '96 The first part of this file outlines LANs in general terms. The second part is about Ethernet which is the proprietary name of the LAN used by BT in the TXE4 Enhancement system. It is explained in pretty simple terms so should be easy and helpful. You may have difficulty trying to decode the ascii diagrams at the end tho ;) ... LANs are commonly used by big (ish) computerised businesses who either buy or hire the LAN (generally from bt), for close office communication. The one popular model is LANStar1, it allows up to 5 users to access the the file server simultaneously using Novell NetWare V3.12 and DOS 6.2X. Introduction:- ^^^^^^^^^^^^ "Local Area Networks" is not a very precise term. To most people the words "Local", "Area" and "Network" conjure up a picture of a network of underground cables and phone lines in an area served by a local exchange. This is not what a LAN is tho... So what is a LAN then? The simplest description of a LAN is that it is a method of providing a communications medium between a number of Microprocessors and / or Microprocessor controlled devices. Apart from the application of a LAN in the TXE4 enviroment, BT is now heavily involved in the provision of LANs for many of its customers. In these cases LANs are used to connect together terminals, computers and other devices in an economic way and allows the user to switch his terminal to the device of his choice. Instead of phone lines, the physical connections are made by different kinds of locally installed cabling. Co-axial cables, multicore cables, twisted pairs and fibre optic cable can be used. Some LANs can be set up cheaply if extension wiring already exists and can be used. The speed of operation of LANs is way faster than the speed that can be achieved on trunk lines and junction network. Analogue circuits with modems are generally limited to about 16k bits. Digital (IDA : Integrated Digital Access) circuits operate at about 80k. LANs on the other hand can work at 10M bits or even faster. These speeds are not merely for connection of the terminals themselves, but rather for the transfer of data from one computer or device to another. Topology:- ^^^^^^^^ Bus System: Within the Bus System there are two distinct and different systems which can be identified: -Baseband transmission system -Broadband transmission system. The essential difference is in the way the bandwidth of the coaxial cable is used for each system. Baseband: Baseband means that the digital signals from the attached devices are applied directly to the cable and therefore employs digital transmission without the use of any carrier or analogue modulation technique. (The signal is still digitally modulated using Manchester phase encoding). This single channel can carry up to 50M bits but only in half duplex mode. Broadband: Broadband, which should have been called "multiband", means that the cable carries two or more high-frequency radio channels. The digital signals are then modulated on to the the radio channels in the same kind of way that TV signals are carried on airwaves. A form of radio receiver on the receiving device separates out the radio frequencies and is able to respond to the digital information on the appropriate channel. Broadband cable has a greater bandwidth which is more fully exploited by frequency division multiplexing. Full duplex operation on one cable is possible by using a frequency converter at the head of the cable to sent the return signal on a different channel from the send signal. Advantages of LANs:- ^^^^^^^^^^^^^^^^^^ "LANs promise a significant increase in performance , efficiency and flexibility for data processing functions." One advantage is that the constituent parts of data processing, central processor, disk storage, printers and terminals, can be purchased in the most efficient numbers to reduce unnecessary duplication. For instance, one printer might be sufficient to serve all users if the communications were possible to each user. Without a LAN, one printer might have been required in each room accommodating terminals. Other advantages (with the appropriate standardisation of systems) are in the flexibility of attaching extra equipment (within limits) to the LAN without affecting existing users or workloads. This equipment would not need to be made by the same company as it would interface directly with the LAN. A further advantage depends to some extent on the topology of the LAN. Distributed control of the LAN is arguably one of the basic features of a LAN since if any essential unit fails the whole system will not fail; the grade of service will be reduced but total disruption of the service should not occur. (Those LANs with central control usually duplicate key units to retain this facility). Ethernet:- ^^^^^^^^ Ethernet is a baseband system which operates at a speed of 10M bits. There is no central controller and all the processors on the network are of equal status. The prime method of transmission is to broadcast the signals along the bus so that each processor (or terminal device) can be either a broadcaster or a receiver of the information. The Ethernet system was the first system of this type, developed initially by Xerox in the early seventies and later by Intel and Digital Equipment Corporation (DEC). Ethernet uses a passive coaxial cable as its transmission medium commonly called its "Ether". Each processor is connected to the Ether through an interface cable to a transceiver which taps into the cable. Diagram 2 shows the topology of Ethernet. The essential features are a passive coaxial as the bus, correctly terminated on its characteristic impedance with stations or other segments tapped to the control bus. When a processor requires a service to be erformed by another processor, it sends a message requesting the service. In this way the processors do not have to be aware of the physical locationof the other processors or the way in which they operate. Control of the bus (the "Ether") is distributed among the processors and each processor contends for the network when it requires to send a packet of information. The possibility of two processors transmitting simultaneously is overcome by using the commonly accepted Carrier Sense Multiple Access technique with Collision Detection (CSMA/CD). Diagram 1 - Ethernet Connections in TXEE ^^^^^^^^^ ________ | | | | |________| | | <- TC ____ ____ __|_ ____ ____ ____ |*TR*|__CC___| |__CC___| |__CC___| |__CC___| |__CC___| | |____| |____| |____| |____| |____| |____| | | | | /\ | | | Tranceiver & ___|___ | ___|___ Connection to | | | | | Coaxial cable | | ___|___ Processor-> | | (100 max) |_______| | | |_______| | | |_______| CC = Co-axial Cable (max length = 500 M) TR = Terminating Resistor TC = Transeiver Cable (max length = 50 M) LAN Messaging:- ^^^^^^^^^^^^^ CSMA/CD Transmission Carrier Sense Multiple Access with Collision Detect transmission is a system whereby each processor wishing to access the bus listens to see if there is a signal on the bus. If so, then the processor waits a while and then tests the bus again. If no signal is detected then the processor proceeds with its transmission. However , the signal takes a finite time to reach it's destination and become established before all the other processors as having control of the bus. During this period another processor could transmit and its signal "collide" with the other transmission. To detect a collision, transmitters listen to the first part of their transmission. If corruption is found, the signal is rejected and both processors "back-off" for a random amount of time and try again. Since the time period of their second attempt is random it is unlikely they will both collide again and one processor manages to control the bus. Communication between sub-systems Diagram 3 shows the subsystems of the TXEE connected to elements of the TXE4 exchange and the two halves of the LAN data highway. The two halves of the LAN consist of coaxial cable and isolator. The LAN isolators are connected together and, in normal operation, join the two halves of the LAN. Message passing over the LAN The messages passed between the subsystems on the LAN are transmitted serially in form of packets. A packet consists of the following:- Diagram 2 - Ethernet Packet Formula ^^^^^^^^^ ________________________________________________________ | | | Preamble : 8 Bytes | |________________________________________________________| | | | Destination : 6 Bytes | |________________________________________________________| | | | Source : 6 Bytes | |________________________________________________________| | | | Type : 2 Bytes | |________________________________________________________| | | | Data : 266 Bytes | |________________________________________________________| | | | Frame Check Sequence : 4 Bytes | |________________________________________________________| Under fault conditions the System Manager Facility (SMF) subsystem isolates the two halves of the LAN. The functions of the subsystem continue, as a Functional Processor (FP) can still access the FP of a replica in the other subsystems connected to the same half of the LAN. Messaging between TXE4 and TXEE The appropriate hardware on the TXE4 exchange translates call events into messages and passes them to the relevant TXEE subsystem. The call event "Register Seize", for example, passes a message to the Set-up Call Handler (SCH) subsystem which includes the following: -calling line identity -calling line type Message processing in the subsystem The subsystem processes the messages in conjunction with data held in files in the subsystems and assembles the result as data contained in messages. The messages are then passed on via the LAN data highway to the designated subsystems on the LAN. Finally - Diagram 3 ^^^^^^^ ^^^^^^^^^ This is a basic (!?) diagram of the Subsystem Communication via the LAN:- _____________________ _____________________ ______________________ | Set-Up Call Handler | | Call Monitoring | | Charge Determination | | Subsystem | | Handler Subsystem | | Facility Subsystem | | _________________ | | _________________ | | __________________ | | | SCH | | | | CMH | | | | CDF | | | | Replica "N" | | | | Replica "N" | | | | Replica "N" | | | | ______ ______ | | | | ______ _______ | | | | ______ ______ | | | || FP A || FP B | | | | || FP A || FP B || | | | | FP A || FP B | | | | ||______||______| | | | ||______||_______|| | | | |______||______| | | | | | | | | | | | | | | | | | | | | | |___|_______|_____| | | |___|_______|_____| | | |____|_______|_____| | | ! ! | | ! ! | | ! ! | | ___!_______!_____ | | ___!_______!_____ | | ____!_______!_____ | | | SCH | | | | CMH | | | | CDF | | | | Replica 1 | | | | Replica 1 | | | | Replica 1 | | | | ______ ______ | | | | ______ ______ | | | | ______ ______ | | | || FP A || FP B | | | | || FP A || FP B | | | | | | FP A || FP B | | | | ||______||______| | | | ||______||______| | | | | |______||______| | | | |___|_______|_____| | | |___|_______|_____| | | |____|_______|_____| | | | ! | ! | | | ! | ! | | | ! | ! | |_____|__!____|__!____| |_____|__!____|__!____| |______|__!____|__!____| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ________ | | | | LAN B | | | | | | | || LAN B | ----+--+----+--+------------+--+----+--+--------+----+--+----+--+|isolator| | | | | | | | |________| | | | | | | | ____|___ | | | LAN A | | | | | | LAN A | | ----+--+--------------------+--+-----------+----+----+--+-----|isolator| | | | |________| | | | | | | | | | ______|____|______ | | | | __|___ | | | | | | FP B |__|__________!______| | | |______| | i | ____|_ | | | | FP A |_________|__________| | |______| | | | | System Manager | |Facility Subsystem| |__________________|