This is a repository for the signal enhancement for magnetic navigation challenge problem (MagNav, and an older version), which was introduced at JuliaCon 2020.
Please refer to "Introduction.pdf" for a brief introduction of our methods and some examples of our results.
Step 1: Copy data files "Flt1002-train.h5" and "Flt1003-train.h5" into the "data" folder.
Step 2: Run "runs/pre_TL.jl".
This pre-filters the data of flight 1003 by the Tolles-Lawson method. A file "runs/data_TL.h5" should be generated by this script.
Step 3: Run "runs/main.m".
The variable "output_line_number" has already been set to "1003.10", which is the line number to be filtered. Two figures demonstraing the performance of the filtering should be generated, and the variable "y_filtered" is the output filtered data.
The Deep Learning Toolbox of Matlab is required.
Ling-Wei Kong, Cheng-Zhen Wang, and Ying-Cheng Lai
School of Electrical, Computer and Energy Engineering
Arizona State University, Tempe, AZ 85287-5706