# NCR24

**Repository Path**: LiuSheng-2020/ncr24

## Basic Information

- **Project Name**: NCR24
- **Description**: Dataset for NCR


- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-04-17
- **Last Updated**: 2024-08-10

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# NCR2024
Dataset for NCR

Context-adaptive navigation, context detection, scenario recognition, integrated navigation, GNSS, GRU, robust navigation context recognition (NCR)

- **data**
At the beginning of each filename, 1 represents open-sky, 2 for tree-lined avenue, 3 for semi-outdoor, 4 for urban canyon, 5 for viaduct-down, 7 for shallow indoor, 8 for deep indoor， and 9 for transition scenario. Note that there is no 6 because we had initially considered a type of scenario to be represented by 6. But after analysis, this scenario type was discarded.


- **ubx2xlsx_parallel.py**. This script extracts features from raw GNSS receiver measurement data (in `ubx` format), including traditional statistical features such as satellites' C/N0 and the number of satellites, as well as the new features proposed in this paper. It also includes two optional functions, `draw_num_cn0()` and `draw_hist()`, for conducting a comprehensive analysis of the satellite features. By default, this program uses a parallel approach to convert multiple `ubx` files into Excel files (for better visualization). The generated Excel files can be read using the [Pandas](https://pandas.pydata.org/) function `pd.read_excel()`.

- **readnmea.py**. Contains various utility Python objects and functions required by the `ubx2xlsx_parallel.py` script. The main functionalities include:
  - Reading NMEA files using the `pynmea2` library;
  - Filtering out unnecessary data, keeping only essential RMC and GSV sentences;
  - Aggregating data over single moments to create a comprehensive dataset.

- **Net.py**. This script includes the deep learning network models used in our study.

- **requirements.txt**. This file lists all the dependencies and packages required to run the provided scripts, including `scipy`, `joblib`, `d2l`, and others. After creating a conda virtual environment, the Python environment can be quickly set up by simply using: `pip install -r requirements.txt`.

  Alternatively, the Python environment can be deployed on a Linux server with a GPU using a Docker container. The Docker image file can be found [here](https://pan.baidu.com/s/1C8jZOGAmQuO0oOyPOXiHFw?pwd=ncr2) (extraction code: `ncr2`).

These codes are now available on our Gitee repository at [https://gitee.com/LiuSheng-2020/ncr24.git](https://gitee.com/LiuSheng-2020/ncr24.git).