Title :
Processor allocation in the star graph using Sg-lattice model
Author :
Wu, Fan ; Hsu, Ching-Chi
Author_Institution :
Dept. of Comput. Sci. & Inf. Eng., Taiwan Univ., Taipei, Taiwan
Abstract :
The star graph has been known as an attractive alternative to the hypercube multiprocessor. When tasks arrive at the star graph, the tasks should be assigned appropriate free processors before execution. A new model, called Star graph (Sg)-lattice, is proposed to model the construction and free configuration of the star graph. Based on this model, the Sg-lattice scheme can fully recognizes the substars. Finally, mathematical analyses and simulation results show that the Sg-lattice scheme outperforms the previous work in storage and time complexities and the average allocation time
Keywords :
computational complexity; multiprocessor interconnection networks; processor scheduling; resource allocation; Sg-lattice model; average allocation time; processor allocation; star graph; storage complexity; time complexity; Analytical models; Computational geometry; Computer science; Fault tolerance; Hypercubes; Lattices; Mathematical analysis; Parallel algorithms; Reflective binary codes;
Conference_Titel :
Parallel Architectures, Algorithms, and Networks, 1997. (I-SPAN '97) Proceedings., Third International Symposium on
Conference_Location :
Taipei
Print_ISBN :
0-8186-8259-6
DOI :
10.1109/ISPAN.1997.645113
