Abstract

Dr. Wen, W. and Xia Xiao.

Scholarship Plan

• (1st year) 09/2022-08/2023: Major/Renowned Funding and Award Schemes (Dr. Hsu Li-ta)
• (2nd year) 09/2023-08/2024: Major/Renowned Funding and Award Schemes (Dr. Hsu Li-ta)
• (3rd year) 09/2024-08/2025: Major/Renowned Funding and Award Schemes (Dr. Hsu Li-ta)

Recent Research Plan: Xia Xiao

Student Supervision

• Supervise Edward? (undergraduate student)

News

• [02/07/2023]
  • extend the conference paper to journal version (JION or GPS solution)
  • integrity (JION)-
    • HPL+ARAIM (solution seperation): part 1
    • Marginalziation: part 2
    • Constrained optimization: part 3
• [12/30/2022]
  • Progress:
  • 1. inject random biases into original measurements, verify the bounding effect over PL of SW+chi-square-test factor
  • 1. compare two implementations of chi-square-test factor: w./o. hinge loss function, observe that chi-square-test factor with hinge loss function will become inactive when WSSE < TD;
  • 1. analyse the Hessian matrix of chi-square-test factor, find that there exist negative eigenvalues, meaning that it’s not convex.
  • 1. case study: when injecting bias into prn 2, 5 and 6; prn 5 is not detected as fault, the possible reason is that there exist healthy satellites (29,61) in the same azimuth, and prn 5 is intended to be estimated healthy too.
  • TODO:
  • increase P_FA, see what happens with decreased T_D;
  • Analyse hessian matrix of pseudorange factors, provide different initinal guess, see under what condition is the hessian matrix not positive definite.
  • 1. Analyse hessian matrix of chi-square-test factor, see under what condition is the hessian          matrix not positive definite.
  • 1. case study: try inject bias into prn 15 instead of prn 5, see if it can be identified as fault.
  • 1. write PLANS conference paper.