Digital Computer Check
Digital Computer Check
the detection of errors during computer operation and the identification of faulty elements and devices. The procedure in solving a problem is classified as preliminary or process control; according to the means used for performing the check, a distinction is made between program (software) and hardware checks.
In a program digital computer check the operation of each assembly is checked by means of testing programs: check routines, to determine the existence of a fault, and diagnostic pro-grams, to find the location of the fault. Diagnostic programs are fed to the input of the equipment being tested, and the output signals obtained are checked against previously compiled tables of serviceable and faulty states of the circuit. A program digital computer check improves the computer’s operating characteristics without the introduction of extra equipment; its principal disadvantage is a reduction in the effective speed of response of the computer.
A hardware digital computer check is usually used as a process control. Most of its methods are based on the introduction of redundancy into the coding of information. Here it is possible to use both the natural redundancy of the codes and artificial redundancy produced by adding separate check symbols or groups of symbols. Digital computer checks according to the absolute value of a prime number are the most common. For this purpose the information being processed is provided with an additional sign, which is the remainder from division of the transmitted number by the assumed absolute value (prime number) or the remainder from division of the sum of the digits in the transmitted number by the prime number. During the reception of the information the check bit is re-formed and compared with the digits transmitted, thus making possible the detection of errors. Correcting codes are used to correct errors during operation. Some of the devices in a digital computer, such as perforators, printers, lighted panel indicators, and power supplies, are unsuitable for hardware checks. Their operation is checked by other methods, such as cyclic checking (inverse operations are performed along with the direct operations, and the data thus obtained are compared with the original values).
Hardware-software checking combines the advantages of the two types of checking; it provides for the timely detection of faults and malfunctions and for quick localization of the trouble.
REFERENCESPutintsev, N. D. Apparatnyi kontrol’ upravliaiushchikh tsifrovykh vychisliternykh mashin. Moscow, 1966.
Mironov, G. A. Ispytitarnye programmy dlia kontrolia elektronnykh tsifrovykh mashin. Moscow, 1964.
Kasatkin, A. S., and A. V. Khruleva. Ratsional’nyi vybor kharakteristik apparatury kontrolia. Moscow, 1970.
G. N. ONYKII