Research On Real-Time Data Transmission Method In Industrial Wireless Sensor Networks

Wireless sensor networks have been increasingly used in the industrial field due to the advantages of low cost,high flexibility,easy maintenance,high density and fault tolerance.Compared with the general wireless sensor networks,the particularity of application in the industrial field makes the data in the network particularly sensitive to the transmission delay.Therefore,the network has higher requirements for reliability and real-time performance,so the design of transmission scheduling method that can reduce network packet loss rate and improve real-time performance has important research significance.In order to reduce the network packet loss rate and reduce the transmission delay,this thesis focuses on the periodic data network and the aperiodic data network,and studies the scheduling method from the perspective of data flow and node.The main contents of this thesis are as follows.(1)The architecture and characteristics of industrial wireless sensor networks are analyzed,and several major related technologies are studied.Especially,the scheduling methods for improving network reliability and real-time performance are deeply studied and compared.(2)Aiming at the network model and working characteristics of periodic data networks,an optimal data flow scheduling algorithm based on backtracking is proposed from the perspective of data flow.The optimal scheduling scheme is obtained by exploring the solution space tree.Then,considering the execution efficiency of the algorithm,a suboptimal real-time scheduling algorithm considering the urgency and concurrency rate is proposed.This method firstly considers the deadline time and remaining hops of the data flow to define the urgency.Under the premise of ensuring the successful transmission of the most urgent data flow,it arranges more data flows to transmit earlier from the perspective of the whole network to improve the overall real-time performance of the network,thereby reduced the packet loss rate and the data flows response time.(3)For the aperiodic data network,a node-based scheduling method is proposed based on the time arrival rate.The method is scheduled according to the superframe.Firstly,the network is layered according to the topology structure.Secondly,the superframe is divided into two parts for the odd number layer nodes and even number layer nodes to send,respectively.And then,according to the conflict relation of nodes in the network and the data density on the nodes,specific time slots are allocated for the nodes.Finally,each node is scheduled specific data transmissions based on the urgency of the data packet.(4)The time performances of the proposed algorithms are analyzed,and the packet loss rate and data transmission delay are compared through simulation experiments.Simulation results show that the methods proposed in this thesis can significantly reduce the network packet loss rate and response time in two scenarios,and improve the network real-time performance.