Design of Self-Checking Discrete Devices Based on Boolean Signals Correction and Composition of Constant-Weight Codes of the “1-Out-Of-4” and “3-Out-Of-4” Types. Part II: The Design Method with Conversion of Some Signals from the Object under Diagnosis

Abstract. To reduce the implementation complexity of self-checking discrete devices in comparison with the design method described in part I of this study, the idea is to consider the composition properties of constant-weight codes of the “1-out-of-4” and “3-out-of-4” types. In this case, it is possible to decrease the number of signals converted in the Boolean signals correction block, fixing this number to p = 4, 3, 2, or 1. Part II presents modified structures for organizing concurrent error-detection (CED) circuits for objects under diagnosis and a circuit design algorithm considering the choice of p. As established below, decreasing p reduces the variability in the choice of codewords belonging to the composition of 1-out-of-4 and 3-out-of-4 codes; this may affect the total checking property of the checker. This drawback can be eliminated by possible permutations of outputs in controlled output subsets and among all outputs of the object under diagnosis. Another solution here is to increase p to 2, 3, or 4. The advantages of the developed method are shown experimentally, and its efficiency is demonstrated in comparison with duplication and other methods based on Boolean signals correction. When converting p = 2 or p = 1 signals in the selected output subset, the gain over the above methods is achieved in almost all benchmarks. According to the experimental results, the developed method has high potential for practical design of self-checking discrete devices.

Keywords: self-checking discrete device, concurrent error-detection (CED) circuit, Boolean signals correction, composition of constant-weight codes, 1-out-of-4 and 3-out-of-4 codes, structural redundancy, controllable structure.