The hypercube interconnection networks is considered as a good parallel architecture for a large class of flexible and powerful networks in large-scale parallel computing. Processors must be allocated to incoming tasks in a way that will maximize the ...
The hypercube interconnection networks is considered as a good parallel architecture for a large class of flexible and powerful networks in large-scale parallel computing. Processors must be allocated to incoming tasks in a way that will maximize the processors utilization and minimize the system fragmentation. In order to achieve this goal, it is necessary to detect the availability of a subcube of required size and merge the released small cubes to form a larger ones. In this paper, we describe the existing strategies for subcube allocation in an n-cube multiprocessors and then propose a new fully-recognizable allocation strategy with [2??+(??)×n] search space in comparison with [(??)×(2??-1)+2??] in binary tree of the TC strategy in generating the search space, while achieving complete subcube recognition. The implementation details of the algorithms are also described.