SPLIT SEQUENCE GENERATION ALGORITHMS FOR EFFICIENT IDENTIFICATION OF OPERATIONAL SUBCUBES IN FAULTY HYPERCUBES

Two new heuristic algorithms are proposed for enhancing operational su bcube identification in hypercubes with faulty nodes. Either of the al gorithms can adaptively construct a complete binary tree based on the fault set such that subcube identification is greatly improved. They c an be usd to modify the standard Gray Code or Buddy strategy to provid e performance improvement in the identification of higher dimensional operational subcubes with practical time complexities that make them a ttractive. In essense, the algorithms provide preprocessing techniques based on recursive splitting of the faulty hypercube so that a large number of high dimensional subcubes can be identified. For the identif ication part, existing schemes can be used to actually harvest the sub cubes.