A unifying construction for difference sets (Q1369732): Difference between revisions

The paper under review clearly ranks as one of the most important papers in the general area of difference sets ever written. I will just mention the most striking results that make it important: (1) It contains the first new parameter family of difference sets found since 1977. (2) It gives a unifying theory which allows to produce a difference set with \((v, n)>1\) in every abelian group which is known to contain such a difference set. Not only was this true at the time of writing the paper, but it also holds for the new families of difference sets which were subsequently discovered by \textit{Yuqing Chen} [``On the existence of abelian Hadamard difference sets and a new family of difference sets'', Finite Fields Appl. 3, 234-256 (1997); and ``A construction of difference sets'', Des. Codes Cryptography 13, 247-250 (1998)]. (3) It characterizes a certain class of groups containing McFarland difference sets. (4) It allows a transparent treatment of the family of Hadamard difference sets, whereas the best previous description (in terms of binary arrays and binary supplementary quadruples) was still rather cumbersome. In particular, the celebrated characterization theorem for those abelian 2-groups which contain a difference set (a problem which took decades to settle) becomes a trivial corollary. (5) It gives a unified way of constructing semiregular relative difference sets in virtually all the known cases, and it gives many new examples.