Stochastic Hybrid System Theory And Method For Information Timeliness On Buffer-Aided WSN

Due to the dynamic change of the environment in which wireless sensor network(WSN)is located,paying attention to the freshness of the received information has become an inevitable choice for such systems.However,traditional network performance measurements are not adequate to characterize the freshness of information.Therefore,researchers have proposed the measurement of age of information(AoI).The traditional AoI analysis method depends on identifying the properties of the AoI sample functions and applying geometric arguments,which often involves convoluted calculations of joint moments.To this end,Roy.D.Yates tried to extend the stochastic hybrid system(SHS)framework and applied it to the queueing theory analysis of AoI.The SHS theory is used to analyze the AoI of mixed queueing systems,which can be modeled by a combination of discrete-state and continuous-time parameters.In summary,aiming at analyzing the AoI of the wireless sensor status update system,combined with specific scenarios such as short packet communication and energy harvesting transmitter,a multi-scenario bufferaided WSN information transmission model is established and analyzed by SHS theory.The main contributions and innovations of this thesis are as follows:(1)To characterize the information freshness in the limited buffer sensor system,this thesis first considers a wireless sensor status update system.It is composed of two sensors and a common destination,where two sensors respectively monitor corresponding physical process.The two sensors,equipped with one unit buffer respectively,will deliver the updates in the form of short blocklength packet with certain probability.Initially,being centered on the combination of each sensor buffer state,the considered system can be modeled as a continuous-time finite-state Markov process.The finite-state represents the different states of the buffers and continuous-time denotes the evolution of the age-related processes.Secondly,the closed-form expression of the average AoI of the system is illustrated by SHS theory.Finally,the simulation results are used to evaluate the performance of the system from the perspectives of signal-to-noise ratio and blocklength.(2)In addition to the bottleneck of information freshness,the performance of WSN is also limited by the limited energy of the transmitter.Thus,a wireless sensor status update system based on AoI is proposed.It consists of two sources,an energy harvesting server and a destination node.Firstly,a three-dimensional state vector could be constructed to track the occupancy of energy buffer,server and data buffer.Then all possible state transitions of the system could be described to establish a multi-dimensional random process of age process evolution.Therefore,the AoI of the system and the moment generating function of AoI could be characterized by SHS,and the system AoI and its higher-order moments can be determined by solving the differential equations.The subsequent numerical simulation results illustrate the importance of considering the second-order and even higher-order moments of AoI in the design and optimization of the proposed system.According to that,an adaptive simulated annealing algorithm is used to study the optimization problem of average AoI under the constraint of AoI maximum variance.This work furnishes theoretical guidance for the design in WSN.(3)In the previous research,SHS theory is used to analyze the AoI of the constructed system.However,SHS theory can only be used to analyze the AoI of the continuous-time status update system and its distribution properties.Moreover,when facing some specific problems,it is difficult to acquire the exact expressions of the state transition equations.Meanwhile,the equipment in an Io T system is normally executed by a time slot.In order to resolve it,a new method for evaluating the AoI of a discrete-time Ber/Geo/1/1 system based on the AoI has been proposed.A two-dimensional state vector is constituted to track the AoI in the destination and the packet age at the server,simultaneously.Consequently,the steady-state probability of each two-dimensional AoI vector can be determined by solving the state transition equations.Afterwards,the distribution properties of the system AoI can be obtained.However,the complex AoI state transition equations make them difficult to acquire the steady state probabilities of the twodimensional AoI vector.Therefore,an AoI assessment method for discrete status update systems based on the probability generating function(PGF)is proposed.In this method,the AoI of the system can be determined by the simple derivative operation on the corresponding PGF,and the AoI’s distribution is settled by writing the corresponding PGF as the power series.The results obtained by this method are compared with those obtained by the previous methods,which further verifies the correctness.Simultaneously,it shows that this method can not only reduce the computational complexity,but also ensure the computational accuracy.