Research Of Method Of Task Scheduling In Internet Of Things

Introduce IoT (Internet of Things) to industry is a main direction that industry upgrade itself to intelligent and informatization. The IoT will be able to gather the information of man and device in production. But there are too much tasks produced by sensors, terminal equipments, user services and so on. The system was expected to deal with this data as quick as possible. There is also many computing nodes in the IoT which can be used to deal with the data. Therefore, how to make full use of the resources in the Internet of things by task scheduling and how to get the best matching relation between the computing resources and the different operating requirements and how to process the data efficiently all were the key problems.Tasks in IoT not only have real-time requirements, but also have partial, communications, and other aspects of constraint, because there are many kinds of tasks in the IoT and tasks having highly combination. In this case the task mapping has been demonstrated for the NP-HARD problem. It means that the optimal solution can hardly be obtained in the required computational complexity. If the tasks pre-scheduling has multi-dimensional constraints, the mapping difficulty will become higher. This is the key problem of static scheduling. When the system is observed from a long time interval, the state of the nodes in the system is not constant, they change dynamicly, which will affect the pre-scheduling of the task. How to control the scheduling in runtime is the key problem of dynamic scheduling. How to correctly combine the advantages of dynamic scheduling and static scheduling is attributed to the hybrid scheduling problem.In view of the above problems, the main research work in this dissertation is:(1) According to the analysis of the research status and the characteristics of IoT, we study and analysis the static scheduling algorithm and dynamic scheduling algorithm.(2) Analyze every constraint in order to design multidimensional constrained optimization objective function for the static scheduling.(3) According to the basic optimization algorithm inconvergence to discrete combinatorial problems, we find the convergence direction in the static scheduling. Base on it, we designed improved Genetic Algorithm to solve the problem of static scheduling.(4) For static scheduling is hard to solve the problem caused by nodes dynamic change, we designed the task retransmission mechanism and node trust degree constrained, than presented hybrid scheduling strategy which has a feedback mechanism.Finally, the related algorithms are simulated. The simulation results show that the improved genetic algorithm is very suitable for the optimization of static scheduling, and its convergence speed is very fast, and the result is very close to the optimal solution. And the hybrid scheduling algorithm can get better results than the static scheduling algorithm when the dynamic state of the node is dynamic change.