Multipoint Terminal Connections

A Data Communication Historical Series

By Bob Pollard

The History of Multi-point Terminal Connections:

   The first terminal to terminal, or a terminal to another device, connection was accomplished using a point to point connection. One device (terminal) was connected to one distant device. As time went by it was possible to connect more than two devices (multi-drop / multi-point) on a line in order to reduce cost and avoid the necessity of constructing individual lines for each of the  connected devices. In the beginning several cities or towns would be connected (multi-point) to a single line and all points (devices) connected to the line would receive all messages sent by any other device. The designated receiver of a message would accept the message and take appropriate action; all other points would ignore the message. Other individual / independent lines from large to medium sized cities would branch out to other multi-point terminal (device) lines. These cities acted as a central relay point and would forward (manually relay) messages from one line to the other appropriate selected terminal line.  This sharing of a line without selection controls is normally referred to as ‘uncontrolled’ terminal operation.

 

   Some simple Morse code selection capability did exist where relays were used on the shared line and the Morse code operator could listen to the messages on the line through an attached line sounder. When a message was addressed (call letters) to a particular office the operator would connect to the line and acknowledge by responding with the office call letters, which indicated the addressed office was ready to receive.

 

   A shared line, in a later developed controlled environment, required that the terminal must be selected before receiving can take place. This avoids the problem, in an uncontrolled situation, where all the terminals receive all the messages being transmitted on the line.

 

   A special mechanical feature built into some of the early Teletypewriters (Teletype / TTY) allowed the TTY to be selected before a message could be received. The TTY used sliding levers (flat bars) and when these levers were placed in the appropriate position the selected TTY would start decoding the message characters and then begin the printing process. The first characters (2 or 3) of a message were the terminal selection characters, and these preceded the actual message content (text). All message and selection characters transmitted on the line were monitored by all the connected terminals, but only the terminal assigned the transmitted call letters would align the sliding levers properly and begin receiving and printing the message text. The selection levers were positioned by an electrical relay mechanism similar to the mechanism used to select and position the proper print key (arm) for printing the character.

 

   Terminal selection functions evolved over time much in the same manner as other communications features and functions. As stated earlier, originally terminals were selected through the sliding bar concept, which was a selection for receiving only. Then more advanced electrical – mechanical (relays) decoding devices were implemented, which allowed terminal selection for both sending and receiving. This device decoded the terminal selection letters and connected the appropriate terminal to the main line. In some cases a different selection code would be used for a send (transmit) connection and a receive connection.

 

   In the 1950s & 1960s multi-point terminal environment, and later, the Terminal control unit, sometimes called the ‘black box’ by maintenance personnel, provided for many functions. Such as: the interface to the MODEM RS-232 signals, controlling the polling and calling recognition (terminal selection), providing parity checking features and provided for other control and protocol functions required for normal terminal operation.

 

   Prior to the 1930’s all multi-point connections were connected to the commonly shared Direct Current (DC) powered line. Both the local terminal environment and the line connected to a distant point utilized DC power, which was provided by batteries or generators. Following the invention of the ‘Carrier device’ and later the MODEM multi-point connections could be accomplished on the DC side of the MODEM or on the AC (alternating Current) side of the line.

 

   Figure 1 is a typical and simplified 1960’s example of a multi-point connection on the AC (analog) side of the line. The ‘Data Set’ was AT&T’s designation for a MODEM.


   The Telephone Company provided this feature at their local exchanges. This allowed multi-point connections across the country anywhere Telephone Company exchanges were located.

 

                                                                    Figure 1

 

   Figure 2 provides a 1970s overview of a multi-point line (network) with device connections on the DC side of the line and a concentrator unit at one end of the network. The example terminals, whether connected directly to the line or through the concentrator could use electrical – mechanical or electronic (control unit) selection devices. The concentrator could be the selected device and in turn control the selection of the connected terminals. The control unit used at the computer or concentrator may be a stand-alone unit or part of the computer or concentrator hardware and may be programmable.

 

                                                                             Figure 2

 

Concentrator:

   A Concentrator does not perform the same functions as a multiplex unit (to be described later); although at times the names are interchangeably used. The normal function of a concentrator is to provide a single line connection to a group of terminals or other devices. The connected terminals appear the same as multi-point terminals connected along a line at different points. When a terminal is connected to a concentrator only one connected terminal can send or receive at a time, which is the difference between the concentrator and a multiplex unit, where more than one terminal can be in operation simultaneously. The problem with the Concentrator designation is the fact Concentrators became programmable and were designed to contain multiplexing and MODEM capabilities, which allow for multiple operational functions. A multiple terminal control (controller) unit in some instances provides the same services as a concentrator. The control unit may be installed in a terminal device, which then will enable other terminals to connect to the line through the terminal with the controller.

 

Terminal Control:

   Figure 3 provides an example of a 1960’s and 70’s terminal selection sequence in a controlled terminal environment. The terminal could be a single terminal on the line or one of several terminals sharing one line (multi-point). Each terminal would be assigned different call characters (terminal address) so only the appropriate terminal would be selected for sending or receiving, all other terminals on the line would ignore the unassigned call/poll characters and would be locked out. The term ‘polling’ is normally used when a terminal is being selected for terminal transmission of a message and ‘calling’ implies the terminal is being selected to receive a message. Numbers and letters identify the sequence of events.

 

  The American Standard Code for Information Interchange (ASCII) is used in these ‘call and poll’ examples since the code provides for acknowledgment and negative responses along with other control characters.

 

                                                                          Figure 3

 

   Figure 4 provides additional details covering the control characters used during a Terminal ‘call’ (terminal receive) sequence (1960s & 70s). Message output is from the computer to the terminal.

 

                                                                        Figure 4

   Figure 5 provides additional details covering the control characters used during a Terminal ‘polling’ (terminal send) sequence (1960s & 70s). Message input is from the terminal to the computer.

 

   Beginning in the 1970’s, concerning terminal selection functions, all that changed was the terminals became programmable mini or microcomputers, line speeds increased and in some cases the selection process was accomplished quicker. Also, since the 1950’s, non-controlled terminals were no longer a problem because the connected communication switching system or computer was capable of servicing the terminals on demand.

 

                                                                        Figure 5

Basic Web page access (Present Day Access):

   Even with the flexibility that was provided with the invention and implementation of the World Wide Web (WWW), usually referred to as the ‘Internet’, which really took off in the 1990’s, not much has changed when it comes to terminal selection. The terminals (Personal Computers) of today still have to be selected before transmitting and receiving can be accomplished.

 

   When a user (client) wants to access a Web page residing on a distant Web server a basic sequence of events or actions will occur. This sequence of events would be initiated and then continue to follow some basic steps, as listed below. A review of Figure 6 may be useful for a pictorial view of these events as they occur. The ‘hash’ marks, - - - - -, represent machines (computers) and the software systems that are variable in number and location.

 

   First, assuming the Web connection is through a local Internet Service Provider (ISP), a connection to the ISP must be accomplished. Using a dial-up line a connection is made using the normal telephone system.

   The user has an URL (Uniform Resource Locator) for a Web page, which is entered in the browser address line. Lets say the URL is http://Autodin.net/alp/mhd.htm and after it is typed in the ‘go’ button or ‘enter’ is pressed. This Web site being accessed contains historical information on the AUTODIN communications system and the above URL will bring up the page containing information on a moving head disk.

 

   The user’s browser connects to the AUTODIN Web server, through the example path illustrated in Figure 6 and the requested page will be displayed on the user’s screen (monitor).

 

The basic steps that occurred behind the scenes to cause this to happen would be as follows:

 

·         The user’s browser broke the URL into three parts:

      1.       The protocol ("http")

      2.       The server name (www.autodin.net)

      3.       The file name (web-server.htm)

 

·         The user’s browser communicates with a ‘name server’, through the ISP, in order to translate the web server name ‘www.autodin.net’ into an IP (Internet Protocol) address, which it uses to connect to the web server machine.

A unique 32 bit IP Address (Internet Protocol Address) is assigned to every computer connected directly to the Internet and this unique identifying number would be a grouping of numbers (up to 12) such as, 192.1.26.2.

·         The user’s browser then connects to the web server at that IP address on port 80.

 

·         Following the HTTP (Hypertext Transfer Protocol) protocol, the user’s browser sends a request to the web server, asking for the file "http://autodin.net/alp/mhd.htm.

 

·         The web server then sends the HTML text for the Web page to the user’s browser. Cookies may also be sent from web server to the user’s browser in the header for the requested page.

 

·         The user’s browser reads the HTML tags and formats the page onto the screen

.

   The page displayed on the screen is now available to be saved or copied for whatever purpose the user may have in mind, although not all web sites allow the page to be directly copied. The displayed page may have links to other pages and, if clicked, will cause the previous step-by-step process to be accomplished for that requested page. A ‘link’ is usually identified by an underline (_____), italics, a different color, named buttons or a list of pages available within the web server. When the mouse arrow is pointed to the link the pointer usually changes to the finger pointing hand.

 

   Also the speed at which the pages are downloaded from the web server to the user’s screen is dictated by the speed of the various communication links between the user and the web site. Usually the slowest bit per second (bps) rate is between the user and the local ISP when a dial up telephone line connection is used. Even though a 56.6 Kbps MODEM is used the actual transfer rate is around 14 to 50 Kbps. This reduction in the bps rate is due to noise (distortion) and electrical interference caused by the line and the various components (equipment) required for a connection between the user and the ISP. This is the reason user’s switch to the broadband digital or cable facilities to increase the bps rate and improve download time.

 

                                                                       Figure 6