Research On Scheduling Technology Based On High Dynamic Data Flow In Wireless Sensor Networks

With the improvement of sensor node capabilities,wireless sensor networks are increasingly being used in communication environments that require high quality of service such as fire detection,intelligent home,medical monitoring,industrial processing and traffic control.In these application scenarios,the existing part of the wireless sensor network adjustment mechanism can not adapt to the environment to change the communication parameters of the nodes,resulting in a low adaptability of the network to complex environments.Although there are some adjustment mechanisms to adjust the communication parameters,due to the slow adjustment rate,it is difficult to meet the service quality requirements in different communication environments.In this thesis,the key technologies of wireless sensor network MAC layer and the common reinforcement learning algorithm are studied.The theory and ideas of reinforcement learning are introduced into the MAC mechanism of wireless sensor networks,and the research questions are transformed into channel access mechanism and retransmission mechanism problem.Aiming at the problem that the traditional MAC protocol is not enough to deal with the complex communication environment,this thesis proposes QoS optimization MAC mechanism based on Q-learning algorithm,QoS optimization MAC mechanism based on distributed round-robin Q-learning algorithm and QoS based on repeated update Q-learning algorithm Optimize MAC mechanism.The innovation point of the algorithm is that the node uses the reliability and delay index in the wireless sensor network communication,introduces the Q-learning learning algorithm,improves the optimization mechanism,improves the self-adjusting ability of the sensor node to the environment,improves the reliability of the network,And reduce the node communication delay.The convergence of Q-learning algorithm is verified,and the feasibility of its application in wireless sensor network communication is verified.Based on the CC2530 physical verification platform,this thesis tests the reliability and communication delay of the three communication protocols and tests the QoS of three communication protocols by changing the topology of the network.And we compare the test results with the original CSMA/CA mechanism and PQ-MAC protocol test results.The test results show that the three QoS-optimized MAC protocols proposed in this thesis have better reliability and delay characteristics than the original CSMA/CA mechanism and the PQ-MAC protocol when the network is stable.After the network is stable,the effective transmission rates of the three protocols are no less than 99%,and the communication delay is no higher than 2ms.