Similarity–Difference Analysis and Matrix Fault Diagnosis of SoC-components
Vladimir I. Hahanov, Mikhail F. Karavay, Vladislav Sergienko, Svetlana V. Chumachenko, Eugenia I. Litvinova, Hanna Khakhanova, Tariq Hama Salih · 2020
Universal metric of data search in cyberspace is proposed; it is based on the use of similarity - difference parameters and matrix structure in binary form. Method for analyzing matrix data structures using the similarity-difference metric for fault detection in digital systems is described. The difference diagnostic method is characterized by the execution of three logical operations on the binary states of the matrix vector-rows and is focused on single and multiple faults in digital systems and software applications. The qubit-difference method for fault detection is characterized by the use of vector parallel logical operations to form a diagnosis of a multivalued technical state, according to the principle of "divide and unite, eliminating contradictions." The advantage of the qubit-vector representation of data in matrix cells is shown, which makes it possible to perform logical operations in parallel in three automatic cycles in order to obtain the required result. Method for detecting single and multiple faults in digital software and hardware systems is described, focused on the hardware implementation of parallel logical register operations, which provide a significant increase in performance compared to existing analogues. Method and hardware implementation of a sequencer for determining similarity-difference-inclusion is proposed, which is characterized by obtaining a more accurate structured assessment of the interaction of two objects.