一种基于服务区分的实时数据传输调度模型研究与实现

Research and Implementation of Real-time Data Transmission Scheduling Model Based on Service Differentiation

面向工业物联网的故障运维数据对传输量、实时性等要求各有不同,在传感信息多样化及大数据量传输条件下,建立高效灵活的通信调度模型,保障事件触发类通信中突发敏感数据的实时传输,是工业物联网无线通信领域亟待解决的重要问题.本文针对当前工业物联网中时延敏感数据传输实时性低,无法根据各类数据不同时延容忍度采取合理调度方法的问题,提出一种基于服务区分的实时数据传输调度模型,将模型动态队列中的报警、状态、图像以及视频四类通信数据分为4种优先级进行传输,利用多优先级感知时隙数据传输方法,将多个时隙组合为一个感知时隙,并根据业务的不同优先级实现感知时隙的合理分配,提高时间敏感数据的实时传输,同时运用模型的良好的动态控制策略,保证缓冲队列的稳定性.仿真实验结果表明,该模型在传感数据量不断变化过程中,队列长度实时调整,时隙资源得到充分利用,保证了高优先级时延敏感数据传输的实时性,有效提高了模型的稳定性.

Fault operation and maintenance data for industrial Internet of things have different requirements on transmission quantity and real-time.Industrial sensor information transmission has the characteristics of diversification and large amount of data.The establishment of efficient and flexible communication scheduling model is an important problem to be solved urgently in the wireless communication field of industrial Internet of things.And it can guarantee the real-time transmission of burst sensitive data in event trigger communication.This paper aims at the problem of low real-time transmission of time-delay sensitive data in the current industrial Internet of things,w hich makes it impossible to adopt reasonable scheduling methods based on different time-delay tolerance of various kinds of data.We propose a real-time data transmission scheduling model based on service differentiation.We divide the communication data of alarm,status,image and video in the dynamic queue of the model into four priority levels for transmission.This model combines multiple time slots into one perceptive time slot by using the multi-priority perceptive time slot data transmission method.And according to the different priority of the business,this method can realize the reasonable allocation of perceived time slots.And we use the model’s good dynamic control strategy to ensure the stability of buffer queue.The simulation results show that the queue length of the model is adjusted in real time and the time slot resources are fully utilized in the process of changing sensor data.This model can ensure the real-time transmission of high-priority time-delay sensitive data and effectively improve the stability of the model.