Assignment of Numbers to Vertices
Assignment of Numbers to Vertices
复制标题
给顶点分配数字
DOI:
10.1080/00029890.1964.11992272
复制
发表时间:
1964
影响因子:
0.5
通讯作者:
J. Lindsey
中科院分区:
文献类型:
--
作者:
J. Lindsey
In a recent paper [1]. L. Harper discussed the following problem. Given a cube in n dimensions, how shall the integers from 0 to 2"-1 be assigned to the vertices of this n-cube in such a way as to minimize the sum, over all neighboring pairs of vertices, of the absolute value of the difference of the numbers assigned to each vertex? He proved that one can do no better than to consider the n-cube as the set of 2" n-tuples of O's and 1's, and assign to the vertex labelled by an n-tuple the integer whose binary expansion is that n-tuple (although there are other assignments that are as good). This problem arose from the following problem in coding theory: assign n-tuples of zero and one to the numbers from 0 to 2"-1 in such a way as to minimize the average absolute numerical error made due to all single errors which arise in transmission of the binary words. Using symbols from a k-letter alphabet instead of from a two letter alphabet, the problem becomes one of minimizing the sum of the absolute values of the differences of the numbers assigned to all pairs of neighboring vertices. Two vertices are called neighboring if they agree in all but one coordinate. Here the numbers from 0 to k"-1 are to be assigned. The proper generalizations of Harper's results are proved in this paper. We shall also reprove Harper's theorem.