A Method of Non-line-of-Sight Measurement and Location

Abstract

The paper aims to solve the wireless Non-Line-of-Sight Measurement and location problem under non-line of sight communication. It introduces modeling the related distance probability and discussing how distance between source and beacons is important throughout the non-line sight compensating and the wireless locating process. The study aims to ensure the accuracy of target location results under LOS/NLOS transmission. The paper opted for an experimental study using the convex analysis method of machine learning theory, including relaxation method and weight least square and approximate substitution transforming the non-convex to convex problem. The approach was complemented by NLOS deviation analysis, including distance measurement model, optimization objective function and constraint condition. The paper provides distance compensation about how measurement distance changes during the practical application situation. It suggests that the non-line of sight compensation positioning method can effectively reduce the location effect caused by different propagation through experimental and practical. The location estimation is more accurate than the other least-squares methods. Because of the hypothesis of experimental conditions, the study results may be idealized. Therefore, researchers are encouraged to test the proposed propositions under different hypothesis. The paper includes implications for the solution of high precision positioning, especially in harsh conditions, the promotion of wireless positioning system. It doesn’t need a number of measurements to identify NLOS. The least squares based related distance probabilities takes advantage of all available measurements and the conditions for occurrence of LOS and NLOS propagation probabilities and the optimization problem can be established by relaxation method.