Sum code formation with minimum total number of undetectable errors in data vectors

In this article the features of sum codes for error detection in data vectors are discussed, that is of interest at the time of concurrent error detection (CED) systems organization. The article describes a classic sum code (Berger code), as well as weighted data bits sum code. It includes the method of code formation, that is optimum with regard to error detection in data vectors. This code has the same redundancy as a well-known Berger code.