With a field replaceable-unit (FRU) (which contains more electronics chips) it. is becoming uneconomical to throw a faulty FRU (module or card or hoard in digital electronic networks) away without trying to repair it (or at least salvage parts).
Testability must therefore also include diagnosability, i.e., the capability for locating faults at least down to smallest-repair-replaceqble-unit.(SRRU) level, in distinction to the FRU. Without diagnosis, repair will be impossible. Many researchers have contributed in diagnostic modeling of digital systems, also the chracterisetion of different diagnostic methods have been established. Most of these works have been concentrated on the identification of faulty units only in digital systems.
In this research, the necessary and sufficient conditions which are required to identify the faulty parts using different diagnostic models, whether these faults in the system units (main units, which are work in normal mode) or in the comparator units (work in test mode) will be presented. Also, an algorithm for distributed diagnosis in computer networks will be proposed.