Cost-effective stochastic resource placement in edge clouds with horizontal and vertical sharing

To support distributed upper-level services such as those in edge or fog computing, a hierarchical and distributed multicloud architecture is introduced to replace the traditional centralized cloud architecture. The intra-time slot demand fluctuation with corresponding multidimensional resource sharing are neglected for efficiency reasons in the existing algorithms. By considering them, the service provider has the potential to obtain better solution, but also makes the scheduling problem a highly complicated stochastic optimization one. To effectively and efficiently solve the problem containing a nonlinear objective and constraints with heterogeneous nonlinear pricing functions, the general stochastic scheduling problem is formulated and its structure is exploited to obtain the base solution quickly, which is further applied with solution closure for fast convergence in the following configurable metaheuristic searching phrase. The experiments with many different problem settings show its superiority over existing algorithms with more than 30% revenue improvement. Therefore, the proposed algorithm can be used as a good supplement to the existing methods.(c) 2022 Elsevier B.V. All rights reserved.