Research On Wireless Communication Transmission Strategy For Real-time Control Systems

With the development and deployment of the fifth-generation mobile communication technology,the future real-time control system needs wireless network to empower it and poses new challenges to the real-time and reliability of transmission to realize the tactile Internet,industrial automation,remote surgery,Smart grid and other emerging applica-tion scenarios.The networked control system(NCS),with its advantages of low deploy-ment and maintenance costs,high flexibility,and strong security,has played a key role in promoting the development of real-time wireless control.However,the development and application of NCS have been severely restricted due to the delay and packet loss in wireless transmission that will affect the control performance.Therefore,how to design NCS reasonably to reduce the impact of delay and packet loss on control performance will become the core issue of this thesis.Aiming at the main design problem of NCS,this thesis proposes combining the relay technology in the packet predictive control system at first to deal with the impact of deep fading based on the existing research results,and exploits the design of NCS using power allocation strategy.Then,considering the effect of status update on system performance,this thesis studies the NCS based on Age of Information(AoI)and optimizes it with static link scheduling strategy.The specific work of this thesis can be summarized as follows:(1)For the real-time network control system where the controller-actuator link is af-fected by deep fading,in order to ensure stability and real-time performance,this thesis proposes a j oint design of packet predictive control and relay technology to achieve NCS,and the total transmit power under restricted conditions,a power allocation strategy is adopted to optimize the system outage probability and control cost.This article first an-alyzes the outage probability of the controller-actuator link,and indicates that the node transmit power and the predicted control length will affect the outage probability and thus the control cost.Therefore,the optimal power allocation with limited total power can be established for the relay system,and the Lagrangian multiplier method can be used to solve it.The simulation results show that this strategy can effectively reduce the system outage probability and control cost,and get the best prediction length as well.(2)For real-time network control systems where the sensor-controller link is affected by deep fading,in order to study the impact of the timeliness of status updates on control performance,this thesis adopts AoI to quantify and optimize the timeliness of the data.First,this thesis analyzes the NCS with estimator,and establishes an optimization problem to minimize the estimation error.Then,the analysis indicates that the estimation error can be characterized as a function of AoI under certain assumptions,thus the optimization problem can be transformed into an optimization problem of the expected AoI.In this system model,the relay technology is also adopted to deal with the effects of deep fading,and a static link scheduling strategy is combined to optimize the expected AoI.Finally,the expected AoI is analyzed through the Discrete-Time Stochastic Hybrid System(DT-SHS)to obtain the optimal link scheduling parameters.The simulation results prove the effectiveness of the expected AoI analysis method and the static link scheduling strategy.