Multipath-aided passive localization using inaccurate receiver based on factor graph

Ganlin Hao, Nan Wu*, Yongliang Li, Hua Wang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

With the growth of ultrawide bandwidth (UWB) wireless network and wearable devices, passive localization with RFID tags becomes more and more important. This paper proposes a multipath-aided algorithm, in which the single reflection signals scattered by the target are used to perform localization estimation. Unlike the existing methods with many accurate receivers, we introduce an algorithm to locate a passive target by a single receiver with position uncertainty. Factor graph representation of the joint posterior distribution of the transmitter and receiver positions is constructed. Taylor series expansion and Kullback-Leibler divergence (KLD) minimization are employed to obtained the proposed parametric belief propagation algorithm on factor graph. Simulation results show that the proposed algorithm is able to estimate the target's position accurately and improve the receiver's position simultaneously with low computational complexity.

Original languageEnglish
Title of host publication2015 IEEE/CIC International Conference on Communications in China, ICCC 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509002436
DOIs
Publication statusPublished - 6 Apr 2016
EventIEEE/CIC International Conference on Communications in China, ICCC 2015 - Shenzhen, China
Duration: 2 Nov 20155 Nov 2015

Publication series

Name2015 IEEE/CIC International Conference on Communications in China, ICCC 2015

Conference

ConferenceIEEE/CIC International Conference on Communications in China, ICCC 2015
Country/TerritoryChina
CityShenzhen
Period2/11/155/11/15

Keywords

  • Factor Graph
  • Inaccurate Receiver
  • KLD
  • Passive Localization

Fingerprint

Dive into the research topics of 'Multipath-aided passive localization using inaccurate receiver based on factor graph'. Together they form a unique fingerprint.

Cite this