3D sensor network location spatial positioning technology based on machine learning

摘要

The purpose of this paper is to combine machinelearning to locate the 3D sensor network space. Real life ismostly a three-dimensional environment. Whether it is afactory in manufacturing or a vegetation base in agriculture,it needs to be monitored and positioned. In this paper,the localization algorithm is discussed to a certain extent.This paper firstly introduces the relevant background andorganizes related work. It also wrote related algorithms,such as ranging-based positioning algorithms in the freespace of wireless sensors. It shows the positioning link byintroducing the wireless sensor network structure systemand node structure. And this paper summarizes theBounding-box Method positioning principle, TDOA algorithmprinciple, and TDOA positioning principle. It thendescribes the gradient boosting tree classification algorithmbased on machine learning, and focuses on the admiralboosting tree classification algorithm related to theexperiment. This paper also describes the ranging technologycombining RSSI algorithm and DV-Hop algorithmin three-dimensional space, and mentions two algorithmsof RSSI and DV-Hop. In the fourth part, the machinelearning coordinate prediction accuracy improvementexperiment and the three-dimensional space positioningalgorithm optimization experiment and result analysis arecarried out. It is proved by experiments that the modelevaluation effect of the gradient boosting tree classificationalgorithm in machine learning is the best. It can be appliedto the calculation of relative position coordinates of labelnodes. It then carried out the three-dimensional positioningeffect test experiment of IDV-Hop algorithm. Thisshows that when the network density in the experimentalenvironment reaches more than 12, the localizationcoverage of IDV-Hop algorithm and DV-Hop algorithm areboth higher than 91. Finally, the hybrid algorithm of RSSIand DV-Hop algorithm is used to compare the positioningaccuracy, positioning coverage and bad node rate withthese two algorithms. It draws the stability of the hybridalgorithm and its effects, and finally discusses and summarizesthe experiments.