Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part II

Research Article

Data Association Based Passive Localization in Complex Multipath Scenario

Download
217 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-73447-7_63,
        author={Bing Zhao and Ganlin Hao},
        title={Data Association Based Passive Localization in Complex Multipath Scenario},
        proceedings={Machine Learning and Intelligent Communications. Second International Conference, MLICOM 2017, Weihai, China, August 5-6, 2017, Proceedings, Part II},
        proceedings_a={MLICOM},
        year={2018},
        month={2},
        keywords={Passive localization Multipath components Data association},
        doi={10.1007/978-3-319-73447-7_63}
    }
    
  • Bing Zhao
    Ganlin Hao
    Year: 2018
    Data Association Based Passive Localization in Complex Multipath Scenario
    MLICOM
    Springer
    DOI: 10.1007/978-3-319-73447-7_63
Bing Zhao1,*, Ganlin Hao1,*
  • 1: Beijing Institute of Technology
*Contact email: zhaobing@bit.edu.cn, hglhust@bit.edu.cn

Abstract

Complex scenarios are characterized by harsh multipath conditions. Recently, strong single reflections among multipath components (MPC) are proved to improve localization performance such as data-association (DA) and multipath components mitigation. We first propose a novel DA method, which figures out the relationship between the received signals and scatters based on an expectation maximization (EM) based Gaussian mixture model. Furthermore, sensors themselves often have uncertainties to be estimated, we propose a joint estimation method to obtain the final estimate. Simulation results show the effectiveness of the algorithm by considering sensors’ uncertainties after demapping. As a result, the proposed algorithm can fit applications of large-scale wireless sensor networks (WSNs) in practice.