Research On Optimization Mechanism Of Internet Of Things Routing

Internet of Things(IoT)and software-defined networks(SDN)have been hot topics in academic research in recent years.There are many heterogeneous networks in the IoT sensing layer,such as wireless sensor networks(WSN)and radio frequency identification networks(RFID).The arrangement of these heterogeneous networks is very complicated.Introducing SDN in IoT can provide individual and flexible management for each layer of IoT.As one of the heterogeneous networks of the Internet of Things perception layer,WSN has been widely used in many fields,such as agriculture,industry,military,environmental monitoring,intelligent transportation and so on.In recent years,WSN has developed rapidly.From the early military field and environmental monitoring fields to the various fields in which people live today,the requirements for reliability and real-time performance of WSN data transmission are continuously improved,so the quality of service(QoS)is provided in WSNs with limited resources.The protection has become one of the most important research issues at present.WSN is a self-organized network of a large number of sensor nodes,often deployed in areas that people cannot reach.Since the WSN sensor node generally uses power supply equipment with limited energy,when the energy exceeds the set threshold,the node will naturally die,and the network topology often changes.Therefore,how to save the energy consumption of the WSN sensing node becomes another difficult problem to solve.This paper first introduces SDN into IoT,and designs an SDN-based IoT architecture(SD_IoT).Secondly,aiming at the optimal routing and energy consumption of WSN under QoS constraints,Multi-path transmission optimization method based on multiple constraints(MMTS)is proposed based on SD_IoT,which includes Link Scoring Algorithms(LSA)and Path Selection Algorithms(PSA).The LSA uses Long-short term memory netwxorks(LSTM),which uses the bandwidth,packet loss rate,delay,and residual energy of the node as input parameters to train the LSTM model and score the link after completing the training.The PSA adopts the Q learning algorithm to select path based on the score and hop count as the return value calculation basis and initialize the Q matrix with the score to realize the selection of the data transmission path in the current environment.