Research On Velocity-assisted Indoor Fusion Localization Algorithms

The traditional wireless localization algorithms assume that the signal is based on line-of-sight?LOS?propagation,but in the actual indoor environment,the signal propagation will reach the receiver through multiple paths,resulting in non-line-of-sight?NLOS?error.Moreover,the complex indoor environment makes it difficult for target nodes to communicate with a sufficient number of reference nodes to complete location estimation.Due to the limited positioning accuracy achieved by a single localization technology,this thesis studies the wireless cooperative localization fusion algorithm based on the distance between nodes and the velocity of nodes under the environment of LOS and NLOS from the perspective of technology integration.Firstly,velocity-assisted second-order cone programming?v-SOCP?wireless localization algorithm in indoor LOS environment is proposed.The objective functions based on time of arrival and received signal strength observations are deduced respectively,and then the second-order cone relaxation technique is utilized to transform them into convex optimization problems.Then,from the point of view of velocity-assisted positioning,the velocity-assisted positioning function is deduced,and the v-SOCP localization estimator is obtained by using the same relaxation technique.The simulation results show that the positioning accuracy of v-SOCP increases by 33.2%when v_max=3.4 m/s and the variance of ranging and velocity is 0.01 and 0.05respectively compared with no velocity assisted.Secondly,in order to make full use of the velocity information of nodes,kalman filter fusion algorithm and adaptive fusion algorithm which based on distance residuals are proposed to further improve the accuracy of positioning system by combining the mobility model with the location estimation of v-SOCP estimator.The simulation results show that,compared with v-SOCP,the positioning accuracy of kalman filter and adaptive fusion algorithm increase by 37.5%and 56%respectively when v_max=3.4 m/s and the variance of ranging and velocity are both 0.05.Finally,two kinds of indoor NLOS wireless location algorithms are proposed.The one is that the constraints are constructed by the discovery that the ranging value is greater than the real distance between nodes under NLOS connection,and the constraints are added to the v-SOCP estimator to restrict the feasible region of the target solution.The second is to optimize the actual ranging matrix by using the position estimated by the velocity-based mobility model,which does not contain NLOS error.Finally,solve it by using the v-SOCP estimator.The simulation results show that both algorithms can significantly mitigate the influence of NLOS error on location estimation,and the two algorithms are suitable for scenarios where NLOS error has less impact and larger impact respectively.This thesis studies the velocity-assisted wireless positioning fusion algorithm.The simulation results show that the velocity information is effective to improve the positioning accuracy,which provides a reference for the research of indoor positioning system.