Research On Wireless Sensor Networks Localization Technology Based On Geometric Constraints

The basic theories and applications of Wireless Sensor Networks(WSNs)have drawn broad attention in recent decades.The popularity of Internet of Things(Io T)technology has also brought great convenience to our lives.Node localization,as the core technology of WSNs and Io T applications,profoundly affects their practicality and effectiveness.However,restricted by energy costs,hardware costs and time costs,localization algorithms adapting to different applications and devices have emerged.This paper starts with the geometric constraint-based algorithm in Static Wireless Sensor Networks(SWSNs),analyzes the significance of geometric constraint to the localization algorithms.Based on that,the geometric constraint is applied to the localization algorithm in Mobile Wireless Sensor Networks(MWSNs),which greatly improves the accuracy of the algorithm.At the same time,the path planning of the anchor nodes,the boundary constraints of the voids and the localizability of the network are studied.The innovations and major contributions of this paper are as follows:Firstly,this paper studies the localization method based on the regional dissection for static networks.By analyzing the Voronoi based localization strategy(VBLS)and error model,an optimal region selection strategy based VBLS(ORSS-VBLS)is proposed.This method takes RSSI measurement error as a constraint.It not only reduces the high failure rate of VBLS,but also uses geometric constraints to further reduce possible error areas and localization errors.Inspired by the VBLS,this paper proposes a Dbelaunay triangulation based localization scheme(DBLS).Compared with the Voronoi diagram,Delaunay triangulation improves the localization accuracy by subdividing a region without increasing computational complexity.The simulation and experimental results show facts as follows:although the ORSS-VBLS consumes more time,it improves localization accuracy greatly;DBLS algorithm reduces localization errors without increasing time consumption.Secondly,the Monte Carlo Localization(MCL)for MWSNs is studied.Inspired by the Voronoi diagram,the Voronoi based Monte Carlo Localization(VMCL)is proposed.Combining the geometric constraints of the region selection strategy,the ORSS-VMCL is proposed.The Kalman filter is used to analyze the real-time measurement error,and the filtering area is further reduced.The Voronoi area in a filtering process is dynamically adjusted.The simulation and experimental results show that VMCL can improve the localization accuracy without increasing the time consumption.The ORSS-VMCL algorithm further improves the accuracy of VMCL,but sacrifices more energy at the same time.Finally,the Moving-baseline Localization(MBL)for MWSNs is analyzed.Considering the application of MBL,a Second-order Moving-Baseline Localization(SMBL)is proposed to improve the adaptability of MBL.The SMBL measures the distance between nodes and better fits real motion of nodes.SMBL reduces MBL's update frequency and saves energy consumption.This article proposes a localization method that combines SMBL and Monte Carlo Boxed localization(MCB),called SMBL-MCB.The algorithm has the advantages of fast operation speed of MCB,and also avoids the problem of high failure rate of MCB when the density of anchor nodes is low or the distribution is uneven.Simulation and experimental results show that SMBL is more suitable for general high-order motion than MBL;SMBLMCB also reduces the failure rate of MCB algorithm.The algorithms proposed in this paper are all based on geometric constraints.Some of the constraints are from RSSI signals and others are from the relative motion of nodes.These geometric constraints play an auxiliary role in the prediction step of the localization algorithms.By adding constraints of the possible locations of the nodes,the localization accuracy can be satisfied.