Performance Analysis And Optimization For Short-packet Industrial Real-time Systems

Industrial Internet is a new infrastructure for the deep integration of new-generation information and communication technology,as well as the manufacturing industry,which motivates the implementation of the manufacturing strategy and the development of the digital economy.The Industrial real-time system is the key technology to support industrial Internet applications such as status monitoring and remote control,with the core of guaranteeing the freshness and low latency of information transmission.Shortpacket communication is one of the key technologies of the fifth generation(5G)of mobile communication.In the industrial real-time system,the data is usually of limited size(e.g.,20-250 bytes)and therefore transmitted with short-packet communication,which is expected to reduce delay and enable real-time sensing and control.Due to the limited coding fault tolerance of short packets and the complex industrial network environment,the analysis of short-packet industrial real-time systems is inaccurate and its resource optimization is not applicable.The specific research challenges are as follows:(1)complex relationship between delay and information freshness due to the finite block length(FBL).(2)low throughput with the single short-packet structure due to complex traffic.(3)low information freshness with a single channel due to the uncertain channel.(4)low resource utilization due to the independent design of communication and control.To address the above challenges,this thesis focuses on the transmission requirements of low delay and high information freshness of short-packet industrial realtime systems,and conducts performance analysis and resource optimization research of short-packet industrial real-time systems.Therefore,this thesis is important to clarify the relationship between various coupling parameters and performance metrics with shortpacket constraints and improve the efficiency of short-packet resource optimization.The main work and contributions of this thesis are summarized as follows.To address the problem of the unclear relationship between delay and information freshness in short-packet industrial real-time systems,age of information(Ao I)is used to quantify information freshness.Also,transmission delay and queueing delay are used to characterize the main system delay,based on which the delay-information age relationship in the FBL regime is investigated.Based on the queueing theory and FBL coding theory,the theoretical expressions for the average(peak)Ao I and delay with respect to block length and update rate are derived,and an approximate relationship between delay and Ao I is derived.The negative correlation between delay and Ao I in the last-come-first-served system with respect to the update rate and their positive correlation with respect to the block length are proved.Based on the derived delay-Ao I relationship,an optimization problem is formulated,where the objective is minimizing the weighted sum of delay and peak Ao I(PAo I).Then,a low-complexity joint optimization algorithm is proposed to derive the closed-form expressions for the optimal update rate and block length.Simulation results show that the proposed algorithm approaches the optimal delay-Ao I performance.The constructed relationship and optimization algorithm provide the theoretical basis for the following research on low delay and high information freshness.To address the problem of insufficient performance due to the single short-packet structure with heterogeneous traffic and complex network environments,the optimization of the short-packet structure in heterogeneous networks with delay constraints is investigated for short-packet industrial real-time systems,based on the derived delay mode.In this section,two packet structures,which respectively adopt independent pilots and shared pilot,are investigated.Also,the delay,reliability and effective throughput are analyzed with respect to the block length and pilot length,where the decoding error and channel estimation error are considered.Block lengths and pilot lengths are jointly optimized for the two packet structures,through instantaneous channel state information(CSI)based dynamic optimization and statistical CSI based static optimization,to strike the tradeoffs among performance,complexity and signaling overhead.It is demonstrated that the independent-pilot packet structure with dynamic optimization is preferable in the scenario with high traffic heterogeneity or high mobility,and that the shared-pilot packet structure with static optimization presents a comparable performance to the former in the case of low mobility,incurring negligible complexity and signaling overhead.The proposed joint optimization algorithm can provide a reference basis for the study of low-delay communication in the FBL regime.To address the problem of low information freshness due to single-channel transmission with uncertain channels in short-packet industrial real-time systems,the optimization of the dual-channel transmission strategy with information age constraints is investigated,based on the derived information age model.Closed-form expressions for the average Ao I and PAo I in the FBL regime are derived with respect to the block length and update rate,based on the discrete-time Markov-chain process.Then,to obtain the explicit region of preference(Ro P)and quantitative PAo I gains of multiplexing/diversity over the singlechannel case,we derive the signal-to-noise ratio(SNR)threshold for transmission mode selection,which is shown to be a decreasing function of the arrival rate and saturates at high arrival rate.Also,the monotonicity of the PAo I gains by multiplexing and diversity,and their achievable gains,are analyzed in a comprehensive manner.It is shown that diversity is able to achieve a PAo I gain of more than 3 d B over the single-channel case at low SNR,while multiplexing has a larger Ro P than diversity and is selected at high SNR and high arrival rate.Furthermore,both throughput and PAo I violation probability are considered alongside the average PAo I for a wide range of tradeoff in system design.The proposed Ao I-oriented dual-channel transmission strategy based on SNR threshold can provide a reference for the study of improving the information freshness in the FBL regime.To address the problem of low resource utilization due to the existing independent design of uplink and downlink in a typical closed-loop scenario of short-packet industrial real-time systems,the joint communication-control optimization of closed-loop real-time systems is investigated,based on the proposed low delay and high information freshness strategies.First,the age of loop(Ao L)is introduced to measure the information freshness of closed-loop real-time systems with two-way delay and error probability.Then,to characterize the average,fluctuating and extreme-case performances,the expressions for the average peak Ao L(PAo L)and its variance and violation probability with respect to the block length and the maximum number of allowable transmissions are derived in the FBL regime.Our analysis shows that the average PAo L is convex with respect to the block lengths and that variance of PAo L and PAo L violation probability are monotonically decreasing with respect to the block lengths.In addition,the coupling between uplink and downlink is quantified and it is proved that the average PAo L is less than the sum of average PAo I in uplink and PAo I in downlink.We then propose a low-complexity PAo Laware block length adaptation scheme and the optimal control,which effectively reduces the PAo L and improves control performance.Simulation results verify our analysis and demonstrate the effectiveness of the proposed policy.