Internet of Things calculation task scheduling system and method

Abstract

The invention discloses an Internet of Things calculation task scheduling system and method. The system comprises a starting module, a task trigger frequency predication module, a task parameter measurement module, a task switching algorithm module, an auxiliary processor module and a main processor module, wherein the starting module is used for stating a node, the main processor module executes former multi-tasks and calculates execution time on executing the former multi-tasks for one time and task trigger situations; the task trigger frequency predication module predicates task trigger frequency for a period of time in future according to the task trigger situations; the task parameter measurement module performs measurement to obtain a plurality of parameters required by the task switching algorithm module according to the former multi-tasks; the task switching algorithm module evaluates whether the optimal execution mode of the next calculation task is a hardware task or a software task according to the task trigger frequency and the plurality of parameters; the auxiliary processor module executes the hardware task; the main processor module executes the software task. By means of respective advantages of the auxiliary processor module and the main processor module, real-time performance of the system is ensured, and overall power consumption of the node is the lowest.

Description

technical field [0001] The present invention relates to the field of the Internet of Things, in particular to an Internet of Things computing task scheduling system and method thereof Background technique [0002] In recent years, the Internet of Things technology has developed rapidly and has been widely used in major accident monitoring, environmental monitoring and protection, intelligent transportation, smart grid, smart medical care and other fields. There is a class of applications with high computational complexity and real-time requirements in IoT applications, such as vehicle recognition and traffic statistics in the field of intelligent transportation. If video sensor nodes are used, image compression / Decompression, pattern recognition and other algorithms. In order to ensure the real-time performance of the system, these complex algorithms usually need to run in a dedicated high-speed coprocessor. Therefore, for such applications, a dual-core architecture is ge...

AI Technical Summary

Problems solved by technology

The coprocessor can quickly switch from the sleep state to the running state at a very low cost; and if it is to switch from the off state to the sleep state or the running state, the coprocessor needs to configure its own program, which will consume more time and energy

Images