Abstract: | With the advent of distributed and multiple processing systems, the study of interconnection networks has developed into a field of its own. In most cases, the performance measures used to characterize the reliability of such networks have been primitive.It is our purpose, in this brief paper, to survey the many available measures of the vulnerability of interconnection networks. The most widely known is classical connectivity; however, measures more recently introduced in the graph theory literature are probably a better practical metric of reliability. Our objective is to make these measures more widely known, and to spur the development of good algorithms to calculate them in practical instances. |