Distributed Subcube Identification Algorithms for Reliable Hypercubes

The Hypercube is one of the most appealing topologies among the interconnection networks designed for parallel processing applications. Most of the basic parallel algorithms, developed to run on an n-cube, can run on a smaller k-cube (k \u3c n ) with a slow-down factor of 2”-k. This partitionability of the hypercube allows algorithms to run on the Operational St&Cubes (OPSCs or sub- cubes with fault-free nodes and links) of an n-cube whose dimensions cannot be fully utilized due to failure of some nodes