The Research On NLOS Localization Algorithm For Wireless Senor Network

Wireless sensor network localization is an important way of positioning,especially plays an important role in the indoor positioning,it is low price,high positioning accuracy;adaptability is used in various fields of people.The positioning accuracy in wireless sensor networks is an important standard to measure the whole positioning system is excellent,In the process of positioning;it needs to deal with the information containing the position parameters measured by the nodes,and locate the labels.so that the labels are positioned and positioned.TDOA-based positioning methods mainly have high positioning accuracy and low system complexity,but require strict clock synchronization in the positioning process.However,in the actual environment often exist in the non-line of sight,need comprehensive consideration of various factors in the process of localization.In this paper,the problems encountered in the development of wireless sensor are introduced,and the solution to the above problems is put forward.Anchor node clock synchronization problem.TDOA-based positioning method has high positioning accuracy,therefore,positioning method used in this paper is the localization method based on TDOA,and proposes a solution of TDOA positioning clock synchronization method based on the.Based on the TOA information received from other anchor nodes.Firstly,the relative clock skew and relative clock offset are estimated.And then the TDOA is compensated by the estimated result,which solves the problem that the time between the anchor nodes is not synchronous and the problem of large positioning error.Location algorithm problem,In the existing location algorithm,Newton iteration,Taylor series method requires an initial value to the location,the selection of initial values have a certain influence on the results of the final estimate,and these algorithms may occur in the iterative process does not converge.In this paper,we propose a two-step weighted least squares algorithm to solve the shortcomings of the positioning algorithm to ensure the positioning accuracy.Non-line-of-sight problems encountered by nodes in actual communication In the process of actual propagation,the signal often encounters an obstacle,which causes the signal to undergo one or more reflections in the process of propagation.The TOA information received by the anchor node has a large error which affects the positioning accuracy.In this paper,we propose a new method to eliminate the non-line-of-sight(NLOS),which divides the anchor nodes involved in the localization into various combinations,calculates the position of the labels for each combination,and finds the residuals of the TOA information from the obtained label positions,Chooses the one with the smallest group from these combinations,the result of this group of data positioning is subject to non-line of sight error minimum or not subject to non-line-of-sight location results.In this paper,the method of suppressing non-line-of-sight is proposed and tested on the hardware platform to test whether the proposed algorithm can achieve the expected results.It can be applied in the actual positioning system.Both the simulation results and the actual results show that the method proposed in this paper has good results and solves the problems encountered in the actual positioning.The proposed algorithm is effective.