الفهرس 
AcknowledgementOnly 14 pages are availabe for public view
 |  | from | 145 |  |  |
| | | from | 145 | | |
Abstract
One of the prime requirements of a computing system is the ability to operate correctly over sufficiently long pe¬riods of time. In the past, computers were used largely in an off-line batch processing node, and, the cons of undetected hardware malfunctions were relatively minor Je now see increasing use of computers in on-line, real-time a-oplications such as the control of chemical process units, nuclear -reactors, and military command and control. Incorrect computer operation in any of these applications can be poten¬tially disastrous. t the Same time, the increasing size and complexity it digital computers have made it more difficult to ensure correct machine operation.
To satisfy these requirements, the system, has to be tested periodically to determine if it functioning properly. If any malfunction is detected, the faulty unit must be identified and re-paired or replaced. The set of tests used should be complete in that it is capable of detecting any fault that is likely to occur, and should also be as small as possible, in order to minimize the time required for this maintenance function. If a fault is found, additional tests then be necessary to identify the faulty unit.