INFORMATION in FAULTY SIGNATURES
Original Publication Date: 1988-Jul-01
Included in the Prior Art Database: 2005-Feb-15
When signature analysis is used and an incorrect signature is observed, it is necessary for diagnostic purposes to identify the failing tests. A method is described for doing this which uses only operations on the signature of the fault-free circuit and the observed incorrect signature. The method is described using a single-input signature analyzer, first assuming a single test failure and then assuming multiple test failures. Lastly, the uncertainty associated with the method when applied to multiple-input signature is described. An n-stage Galois Field divider over GF(2) is used as a single- input signature analyzer to compress the m bits of the input data stream Z into an n-bit signature. The characteristic polynomial of the divider is: f(x) = 1 + h1x + h2x2 +...