科研成果详情

We propose a new criterion "Entropy based on projection Hamming distance pattern", denoted as S(D), to evaluate and detect combinatorially non-isomorphic orthogonal designs. It is defined by the entropy based on the distribution of all Hamming distance patterns of projections, that provides a new perspective to evaluate orthogonal designs. Hamming distance pattern is incorporated in this new criterion since it has many connections with other criteria in the literature, such as the generalized word-length pattern and uniformity measures. To compare the detection performances, we adopt three orthogonal designs with different sizes, 18-run, 27-run and 32-run. Some non-isomorphic classes of 27-run and 32-run orthogonal designs are misclassified by S(D). Hence we adopt the projection Average Squared Correlation distribution as an amendment in the detection algorithm to improve the effectiveness.