Host-Controlled Fault Diagnosis in a Data Communication System

The present invention relates to data communication systems comprised of one or more host computers, one or more terminals which communicate with a host computer, one or more communication circuit elements supplied by a common carrier tariffed by the Federal Communications Commission (FCC) and other circuit elements supplied by nontariffed vendors, such as modems, compactors and multiplexors which contribute to the correct and efficient operation of the system. In particular, the invention relates to the rapid determination of which component or circuit element has failed in the event of a system fault.

Present-day society is becoming increasingly dependent upon data communication systems. In the event of the failure of such a system, workers and facilities frequently stand idle for considerable periods of time. Rapid restoration of service is thus a matter of considerable economic significance. Determining which particular component or circuit element of the system has failed is typically done by an expert skilled in the design and operation of one or more circuit elements of the system, and is frequently much more time-consuming and error-prone than is desirable. In practice, the difficulty of prompt diagnosis of system faults is compounded by the fact that the circuit elements comprising most data communication systems are supplied by different manufacturers and vendors. If the wrong vendor is mistakenly called on to provide repair service, the actual restoration of service may be delayed by hours or days. As data communication systems have become increasingly complex, a systems analyst becomes less and less able to rapidly and accurately diagnose the cause of a system failure.

The object of the invention is the provision of a novel method and system for rapidly and accurately diagnosing the cause of failure of a data communication system.

In the transfer of digital data between a host computer and a remote data terminal via a data communication system that includes one or more circuit elements supplied by a common carrier tariffed by the FCC, the present invention provides automatic diagnosing of faults within the system. It preferably comprises a series of digital logic loopback networks, each individually connected in parallel between the transmit and receive portions of the data communication system at the terminals of the various data handling circuit elements of the communication system. Each logic network operates in selective and independent fashion in response to digital addressing signals sent by the properly programmed host computer; thereafter other digital signal patterns sent from the host computer enable the computer itself to test the integrity of data transmission through each circuit element of the communication system by looping back those digital signals to the computer. Each set of digital addressing signals can be uniquely associated with one and only one of the circuit elements.

Other advantages, objects and attributes of the invention will become more apparent from a description of an embodiment thereof in which the following Figures appear, to wit:

FIG. 1 is a schematic diagram of a data communication system in which the present invention is particularly useful;

FIG. 2 is a detailed schematic diagram of a prototype of the present invention; and

FIG. 3 is a program listing of the host computer program which was used in conjunction with the operations of the present invention.

Refer now to FIG. 1, in which a host computer 10 is shown connected to a data terminal 11 via a series of circuit elements, viz., a pair of low-speed modems 12, a pair of multiplexor units 13 (which typically have more than the single low-speed input here shown), a pair of high-speed modems 14, and a telephone line 16. It may be appreciated that rapid and accurate diagnosis of a system failure is important. Experience has shown in accordance with the present invention that such diagnosis can be provided quickly, accurately and rapidly by installing a series of loopback networks 20 of the present invention at points A, B, C, D, E and F, as shown; i.e., at the output terminals of each of the circuit elements. When each network 20 is energized to locate a fault, the host computer 10 first signals the diagnostic network 20 of interest, say at A, to assume a loopback mode, as explained below; thereafter other digital signals -- in a predetermined pattern -- are sent from the host computer; those system circuit elements more remote from the host computer are left isolated; and the data pattern that is rerouted back to the host by network 20 at Point A is checked against itself for accuracy at the computer. After the network 20 at A is inactivated by the host taking it out of the loopback mode, the host places a similar network, say at B, into a like loopback mode, and the process repeated. By proceeding this way, the host may rapidly and accurately determine between which two loopback networks the fault lies.

FIG. 2 illustrates a diagnostic loopback network 20 in detail.

The operations code from the host computer initiates operation of the decoder network 20 as follows: it first energizes integrated circuit (IC) chip 30, which converts electrical levels of the signal from the EIA standard to those required by the following transistor-transistor logic (TTL) elements. The converted operational signal first activates IC chip 31, a Universal Asynchronous Receiver-Transmitter (UART), whose main task is to internally store the data bits which make up the operations code as they are presented serially by IC chip 30, and then finally present all 8 bits in parallel to comparator chips 33. IC chip 34 is a D-type flip-flop whose function is to disable the comparison operation while the output of IC chip 31 is in the process of changing state. Clock circuit 35 is standard, and is needed for the proper operation of IC chips 31 and 34.

IC chips 33 are comparators that compare the operations code which was received to a predetermined bit configuration, in this case 0001 1110. When this particular operations code is received, a pulse is sent to IC chip 36, a D-type flip-flop. When the output, Q, from IC chip 36 is high, then the diagnostic network 20 is in the loopback mode. When the output Q is low, then the diagnostic network 20 is in the normal mode and the output Q is high. Each time the special operations code is received, the diagnostic network, and in particular flip-flop 36, flips from the one mode to the other mode. Manual reset button 40 can be pressed at any time. The diagnostic network 20 is then forced into the normal mode at that time.

The Q and Q outputs from IC chip 36 are used to control the operation of the three AND gates 41, 42 and 43, comprising, along with OR gate 44, a gate network 45 that controls signal flow to and through each diagnostic network 20. Specifically, when in normal mode, gate 41 is disabled while gates 42 and 43 are enabled. Thus, signals which enter through level converting chip 30 exit through level converting chip 50, to data terminal 11, FIG. 1; signals which enter via level converting chip 51 exit via level converting chip 52. On the other hand, when in loopback mode, gate 41 is enabled while gates 42 and 43 are disabled. Thus, signals passing through IC chip 30 exit via IC chip 52; no signals pass through IC chip 50; similarly, signals which enter through IC chip 51 are not transmitted past gate 43. The OR gate 44 accepts signals from either gate 41 or gate 43, which signals then pass to IC chip 52.

FIG. 3 illustrates a computer program which has been executed in the properly programmed host computer 10 of FIG. 1 to test the operation of the diagnostic network 20 of FIG. 2. In this program, the special operations code which causes change of mode in a single diagnostic network 20 consists of one character, and is called "TOGGLE." The program transmits four characters, "ABCD," followed by the single character "TOGGLE," followed by a second group of four characters "EFGH," followed by the operations code word "TOGGLE," followed by a third character set "IJKL." Proper operation is demonstrated by the characters "ABCD" proceeding to the terminal beyond the loopback network 20, the characters "EFGH" looping back to the host computer, and characters "IJKL" again proceeding to the terminal.

In this case, the host computer is a IBM 370/68.

Note that in FIG. 3 the command "TOGGLE" in effect flips the network 20 of FIG. 2 from one mode to the other mode independent of its previous state. Thus, assuming the network 20 started in the normal state, the first-mentioned operations code command (i.e., "TOGGLE") of FIG. 3 places the network 20 in a loopback mode while the second command returns the system to its normal operating mode.

The invention is not limited to the above combinations of elements and steps, inasmuch as substitutions are readily apparent to those skilled in the art. For example, synchronous as well as asynchronous data protocols might be employed within the data communication system to which this invention relates. Bisynchronous (BSC) or Synchronous Data Link Control (SDLC) are examples, and may enable the host computer to operate in a similar but slightly different manner than previously discussed. Such protocols may require that each loopback network engage in a predetermined dialog with the host computer in order to establish control by the latter. Also, the loopback network itself may be packaged as part of a data handling element in the data communication system, such as within a modem, concentrator, compactor, multiplexor, data service unit, or terminal. Also, each diagnostic network, upon command by the host computer, may perform a diagnostic logic function different from loopback. For example, a test signal could be generated upon interrogation, and that signal measured by other equipment controlled by the host computer; moreover, the network itself may take measurements and send the results back to the host computer for analysis; or it may loop back the analog circuit rather than a digital interface, as discussed above. Thus, the invention is to be given the broadest interpretation within the terms of the following claims .