A global optimization approach for three layers of computational grid stack (Q1028178)

Grid architecture is organized into five layers: fabric layer, connectivity layer, resource layer, collective layer and application layer. The authors consider the requirements of fabric layer, collective layer and application layer at the same time, and propose a global optimization approach for three layers of grid stack. The global optimization for three layers of grid stack can be decomposed into three sub-problems: resource allocation for fabric layer, service integration collective layer, and application quality of service optimization problem at application layer. The proposed algorithm decomposes the global grid optimization problem into a sequence of three sub-problems at three layers through an iterative algorithm. The convergence of the iterative algorithm is proved. Simulation studies of the proposed algorithm show that the proposed algorithm is more efficient than the single layer optimization when quality of service requirements are involved in multiple layers of the grid architecture.