Loading patent…
AssigneeTexas Instruments Incorporated
InventorsPeter Sautter, Harald Tschentscher, Carsten Gebauer, Berthold Fehrenbacher, Roy M. Haley, Alexandre Palus, Charles Ming-Fong Tsai, Venkata Kishore Gadde, Hoi-Man Low
Estimating the term…
In order to detect a faulty error correcting unit (2) in an embedded system, wherein the error correcting unit (2) receives output data from a data source (20) and determines, whether the received data are incorrect, and wherein if the received data are incorrect, the error correcting unit (2) is expected to correct at least one error within the received data, output the corrected data and manipulate an error vector (4), a method and a system (Ia) are suggested that enable to compare the output data of the error correcting unit (2) with at least one reference data, wherein the at least one reference data originate at least indirectly from the data source (20). Both, the error vector (4) and the result of the comparison are input to a plausibility test in order to decide, whether the error correcting unit (2) is faulty. According to the result of the plausibility test, a failure vector (7) is manipulated in order to indicate whether a failure in the error correcting unit (2) is detected.
In a method for detecting a failure in an error correcting unit (2), wherein the error correcting unit (2) receives output data from a data source (20) and determines whether the received data are incorrect, and wherein if the received data are incorrect, the error correcting unit (2) is expected to correct at least one error within the received data, output the corrected data, and manipulate an error vector (4), wherein the error vector (4) at least indicates whether an error was corrected or whether an error was identified but could not be corrected, the improvement comprises the steps of
An apparatus (1a, 1b, 1c) for detecting a failure comprising
(canceled)
Dependencies are inferred from the claim text.