Research On Scheduling Strategy Of Internet Of Things Terminal Based On Task Relevance

The Internet of things(IOT)is considered to be the third revolution of the information technology industry,and is going to be widely used in many fields.The IOT terminal is of great importance for its capability of perceiving information in the physical world and exchanging the information between “things and things” and“things and humans”.However,the cost and power consumption of IOT terminal limit its computing power and storage capacity,and further decrease the efficiency of task execution and response time.In order to adequately schedule the limited computing and storage resources and further improving the operating efficiency,considering the existing relevance among multiple tasks in the IOT terminal,this thesis proposes a scheduling strategy based on the task relevance.The strategy consists of two parts: task scheduling and storage resource scheduling.The main studies are as follows:1.Aiming at the complicated relevance among IOT terminal tasks,this thesis proposes a task scheduling strategy based on task relevance.A task model with the job polling group constitute the main body of this strategy and a priority factor matrix is established as the basis for task scheduling according to the task execution time.The priority incremental matrix parameterized by the task relevance is generated to dynamically change the task priority for the real-time of predecessor task after each task is executed.For aperiodic task,this strategy constructs temporary job polling groups for preemptive scheduling.The experimental results show that this scheduling strategy can effectively increase the throughput of tasks and shorten the response time of aperiodic task.2.To solve the problem of memory resource shortage in IOT terminals,this thesis proposes a storage resource scheduling strategy based on task relevance.The entropy weight method is used to calculate the weight of task migration and the priority of data segment loading is set according to the number of the tasks which is using the data segment.When the memory usage exceeds the threshold,the data segment is selected to be moved out of memory to external memory according to the task migration weight and load priority to ensure the memory requirements of the tasks with strong correlation can be met preferentially.The experimental results show that this strategy effectively reduce the influence of memory overload on task throughput and aperiodictask response time.In summary,starting from the relevance among tasks in the IOT terminals,this thesis studies task scheduling and storage resource scheduling,and established simulation and test platform to test and verify the performance respectively.The test results show that the scheduling strategy proposed in this thesis effectively improve the task throughput and shorten the aperiodic task response time.