The DMS-100 is a member of the Digital Multiplex System (DMS) product line of telephone exchange switches manufactured by Northern Telecom. Designed during the 1970s and released in 1979, it can serve up to 100,000 telephone lines. The purpose of the DMS-100 Switch is to provide local telephone service and connections to the PSTN. It is designed to deliver services over subscribers' telephone lines and trunks. It provides plain old telephone service (POTS), mobility management for cellular phone systems, sophisticated business services such as automatic call distribution (ACD), Integrated Services Digital Network (ISDN), and Meridian Digital Centrex (MDC), formerly called Integrated Business Network (IBN). It also provides Intelligent Network functions (AIN, CS1-R, ETSI INAP). It is used in many countries throughout the world. There are also DMS-200 and DMS-250 variants for tandem switches. Much of the hardware used in the DMS-100, with the possible exception of the line cards, is used in other members of the DMS family, including the DMS-200 toll switch.
Hardware All power distribution is at -48 VDC (nominal), from which DC to DC converters on every shelf provide other necessary voltages.
Central Control Complex (CCC) The Central Control Complex comprises the Central Processing Unit (CPU), Program Store (PS), Data Store (DS) and the Central Message Controller (CMC).
The CPU contains two identical 16-bit processors running in hot standby mode. The original CPU core was referred to as the NT40 CPU and was implemented in approximately 250 discrete logic devices across several circuit boards running at 36 MHz. The NT40 core consisted mainly of the NT1X44 stack card, which provides some register and stack functions of the processor, the NT1X45 which contained the arithmetic and logic functions, the NT1X46 which provides more registers and the load-route read-only memory (ROM) and the NT1X47 timing and control card which provides the micro-cycle source and microstore decoding functions of the processor. The NT1X47 card also contained the 2-digit hexadecimal display to indicate test result codes and the condition of the core. The NT1X48 processor maintenance card contained a thumbwheel on the faceplate to enable various diagnostic tests of the CPU. A later modification of these same five circuit boards with faster pin-compatible discrete logic devices enabled the CPU to operate at 40 MHz allowing central offices to improve call throughput capacity by 10 percent. When the CPU is configured in dual hot standby mode, a mate exchange bus (MEB) between the two CPUs enables the state of one CPU to be continuously compared to that of the other CPU on a cycle by cycle basis. Any discrepancy between the two CPUs results in maintenance circuitry determining which CPU is at fault and activity to change to the same CPU. A Program Store is dedicated to each CPU and is a memory for the program instructions required by that CPU for processing calls, maintenance and for administrative tasks. The PS associated with the other CPU contains identical program instructions. A Data Store is dedicated with each CPU and contains dynamic information on a per-call basis, as well as customer data and office-specific settings. The other CPU is also associated with its own DS containing duplicate data. The Central Message Controller controls the flow of messages between the other units of the CCC and prioritizes them for the Network Message Controller (NMC) in the various Network Modules (NM), or the Input/Output Controller (IOC). Both CPUs have access to either CMC which share the message load to the Line Modules or peripherals. The original NT40 based CCC was replaced by the compatible DMS SuperNode in 1987.
DMS SuperNode The DMS SuperNode Computing Module was first based on the Motorola 68020 Central Processing Unit (CPU) and then upgraded to the Motorola 68030. In the early 1990s it was further upgraded to use the Motorola 88100 and 88110 Reduced Instruction Set Computing (RISC) CPUs. This RISC version of the SuperNode Computing Module was known as the BRISC (BNR Reduced Instruction Set Computing) CPU. With the BRISC CPU the DMS SuperNode had a processing capacity of 1,500,000 call attempts per hour. DMS SuperNode featured increased processing capacity across a distributed architecture allowing for the development of new features and services. Each of the elements of the DMS SuperNode uses a common SuperNode CPU hardware design differing only in the software used to control them. The SuperNode consists of two main elements: DMS Core and DMS Bus.
DMS Core provides the main computing facility and is made up of the Compute Module, System Load Module and a Message Controller. The Compute Module contains redundant SuperNode CPUs to handle call processing and maintenance functions and, like the NT40 core, can operate in a synchronized mode with its mate. The System Load Module contains all the necessary software for every element of the DMS switch and also provides file system and data storage functions on magnetic tape and hard disk. The Message Controller provides communications links between the DMS Core and the DMS Bus. DMS Bus is used to interconnect the DMS Core, the switching network and the Input/Output controller (IOC) and manage message flows between these units and consists of redundant Message Switches. The Message Switches of the DMS Bus operate in a load-sharing mode and one of them provides the main clock source for the DMS-100 system while the others are synchronized to it. Messages between all SuperNode units are carried by optical DS512 links. The operating system used by both generations of the DMS-100 switch was called Support Operating System (SOS) and was written in a high level language called PROTEL which stood for PRocedure Oriented Type Enforcing Language developed at Bell Northern Research (BNR). Hardware and maintenance are administered locally through cathode-ray terminals, through a multilevel menu system called MAPCI. There are various methods used to access the DMS remotely as well, including modem and telnet. Backups and other hard drive work are administered through a DISKUT command line program.
… excerpt ends here. Continue reading the full article.





