Towards scheduling for Internet-of-Things applications on clouds: a simulated annealing approach

摘要

The Internet of Things (IoT) and Cloud Computing are two novel paradigms both rapidly evolving in theirrnparticular areas of application. The former is enabled through several technologies ranging from communicationrnsystems to distributed intelligence, whereas the latter provides the means for massively parallelrncomputation on demand. Therefore, we can include Cloud Computing as an enabling factor in the greaterrnpicture of the IoT. Given the complexity of IoT’s computing concepts, it would be prudent to assume thatrnthis complexity will eventually provide computing workloads that are different in kind to the current ones.rnThus, it becomes important to study how current scheduling optimization techniques can be adapted to suchrncomputing tasks. In this study, we have evaluated the application of simulated annealing in a multi-cloudrnsystem serving a workload of processes with low parallelism but with high arrival rates and highly variantrnrun-times. A discrete event simulator was used in order to assess both the performance and the cost of thernsystem. Simulation results indicate that significant gains both in performance and in cost can be achievedrnusing the proposed scheduling technique in this context.