Low Latency And High Reliability Wireless Resource Management For Industrial Automation

Real-time wireless feedback control completes assigned tasks through driving automation and is considered as a core application in industrial automation scenarios.In such systems,the goal of wireless communication is not only to improve reliable network services,but more importantly,to ensure that the critical information needed by the controller can be updated in a timely and effective manner to support the information flow of the entire automation control.However,traditional communication designs often start only from the perspective of wireless communication,leading to unnecessary network resource overhead.In some cases,congested networks can even slow down the update of some critical information,leading to bad control performance.This thesis intends to start from the perspective of control and discuss how to jointly optimize the parameters of both the communication system and the control system to achieve customized wireless communication services for industrial automation.This thesis first investigate the PPC(Packetized Predictive Control)of how the prediction length affects communication requirement,with which a co-design for communication and control parameters is proposed.Then,this thesis focuses on real-time feedback wireless control systems,discussing the impact of information freshness on control performance and proposing communication indicators for control stability and convergence speed.Based on this,a new wireless information update strategy is designed.Subsequently,the design scheme is extended to practical cases,starting from the perspective of real-time control of robotic arms,designing a prototype,and verifying the proposed algorithm on top of it.Finally,the thesis discusses how the control system affects the channel fading by influencing the spatial position of mobile devices and designs an algorithm to estimate the average channel capacity.The specific work of this thesis is as follows:(1)In PPC systems,an algorithm for joint optimization of wireless communication time-frequency resources and control prediction length has been designed.Specifically,when the control command costs small size of communication payload,and requires extremely high reliability on wireless communication,the thesis finds that the wireless resource overhead decreases as the communication throughput increases,meaning that maximizing communication performance does not always optimize the control results.This conclusion validates the core viewpoint of this thesis,which is that customized communication services should be designed from the perspective of control performance.(2)The impact of information freshness on control performance is discussed separately under two control systems.In the asymptotically stable control system,stability is the core focus of the entire control device.Therefore,the age of stale information(Ao SI)is proposed to link the control system stability to communication designs,and ultimately,maximizing Ao SI achieves statistically optimal control stability.In the completely stable control system,the control system becomes insensitivity to information freshness.Meanwhile,introducing appropriate latency to improve control convergence rate can actually enhance control performance.Therefore,this thesis also designs another indicator called value of information(Vo I)to link communication parameters to the control convergence rate.Finally,by maximizing the control convergence rate,both control and communication performance improvements are achieved.(3)Considered a general control system,the impact of communication on control is further discussed.Specifically,this thesis utilizes the information bottleneck principle to divide the source information as two parts: Feature and noise.Then,the feature is used to quantify the information importance.Meanwhile,this thesis proves that only when the noise is independent with the feature,and keeps stable during the control process,the traditional method for optimizing increment of information is equivalent to the proposed method for optimizing increment of feature.Finally,a prototype for robotic arm control is designed based on this framework,and the designed strategy is verified on it.(4)The study investigates how the state of the control system can inversely affect wireless communication.We consider a typical wireless communication scenario between a drone and a base station,where the drone’s flight trajectory affects the communication’s channel fading.By modeling the control system of the drone,this thesis obtain the expression of the misalignment error.Finally,the average capacity during the control period in terms of the misalignment fading is obtained.The results of this thesis can significantly reduce communication overhead while ensuring control performance.More importantly,the designed framework has been validated through a prototype,which has practical significance and provides new ideas for the joint design of future industrial automation.