Single frequency network based mobile tracking in NLOS environments

Jun Yan*, Kegen Yu, Lenan Wu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

In single frequency network (SFN) positioning, base station (BS) identification is inevitable and non-line-of-sight (NLOS) propagation is usually dominant especially for indoor scenarios. BS identification and NLOS mitigation are two challenging problems which have significant impact on the SFN positioning performance. In this paper, a mobile tracking scheme is proposed to deal with these challenging issues. Specifically, BS identification is first formulated as a data validation problem. Each time-of-arrival (TOA) measurement is tentatively associated with a specific BS so that a number of TOA-BS relationship sets are produced. The gate technique is adapted to evaluate all the TOA-BS relationship sets and the set with the smallest gate parameter value is selected. This identification technique is suited for both line-of-sight (LOS) and NLOS propagation scenarios. The interacting multiple model (IMM) smoother is then utilized to smooth the identified TOA measurements at each BS to reduce the NLOS errors. In addition, the position determination and BS identification are jointly considered to enhance position estimation accuracy. Simulation results demonstrate that the proposed SFN positioning approach can perform satisfactorily in different propagation scenarios and has better performance than other SFN positioning algorithms.

Original languageEnglish
Pages (from-to)54-67
Number of pages14
JournalPhysical Communication
Volume13
Issue numberPA
DOIs
Publication statusPublished - 1 Dec 2014
Externally publishedYes

Keywords

  • Base station identification
  • Digital TV signals
  • Mobile tracking
  • Non-line-of-sight
  • Single frequency network

Fingerprint

Dive into the research topics of 'Single frequency network based mobile tracking in NLOS environments'. Together they form a unique fingerprint.

Cite this