Probabilistic analysis of memory reconfiguration in the presence of coupling faults

Abstract

The problem of reconfiguring memory arrays using spare rows and spare columns has received a great deal of attention in recent years. However, most of the existing research assumes that the array contains only stuck-at faults. This paper, addresses the problem of reconfiguring memory arrays containing both stuck-at faults and coupling faults. The authors present a probabilistic model for studying this problem which is known to be NP-complete. In this model, they distinguish between two classes of faults, namely the class of stuck-at faults and the class of coupling faults. All faulty cells in an array are assumed to independently distributed. The authors first present a bound on the probabilities of occurrence of these two classes of faults that will allow almost all problem instances to be reparable. They also present a bound on these probabilities of defects that will make reconfiguration almost impossible. Empirical study is carried out to validate theoretical results and to investigate the nature of problem instances with probabilities of defects that do not fall within the theoretical bounds.<>