Research And Implementation Of Task Scheduling Mechanism Based On Service Platform Of Internet Of Things

Ubiquitous Service Platform for IOT (Internet of Things) combines originally isolated Internetof Things business, provides comprehensive business support and services, and applies a variety ofapplications to share resources and information, which greatly promotes the development of IOT.On the one hand, large numbers of tasks from different applications have different requirements forreal-time and critical performance. The classic real-time scheduling algorithm, which takes only oneparameter into account, is difficult to meet scheduling requirements in terms of reasonability andefficiency. On the other hand, an open and complex operating environment needs the taskscheduling mechanism not only to achieve a reasonable task scheduling, but also to maintain systemstability as much as possible.Based on the analysis of the classic real-time task scheduling algorithm and the imprecisecomputation task scheduling model, this paper puts forward a hierarchical task schedulingmechanism suited for Ubiquitous Service Platform of IOT. In the upper scheduling algorithm, wepresents a dynamic feedback elastic task scheduling mechanism, which provides a dynamicfeedback on the number of tasks missing and system resource utilization, and the scheduling get thecurrent running state and change the elastic task cycle to adjust the system load and maintain thestability of the system in real time. At the same time, the underlying scheduling adopts IVDAscheduling Algorithm (Integrated Value Deadline Algorithm). The IVDA algorithm computes thepriority of the task according to the task deadline and the task value, through adjusting the priorityweighting parameters to adapt to different application environments. Finally the system obtains abalance between real-time and critical.In the existing services platform of IOT, through analyzing the ability of convergence from theunstable state to the optimum work state, we verify the effect of dynamic feedback schedulingmechanism for maintaining the stability of the system. In comparison with other algorithms in termsof hit value ratio and differentiated guarantee ratio, we prove that the IVDA algorithm caneffectively trade off between the real time performance and the priority of the tasks in IOT..