gnss-benchmark 2.3.0

GNSS benchmarking tool

GNSS processing engine benchmarking data set

This repository contains several datasets to compute the performance of a GNSS processing engine. This repository has been developed to help Rokubun Jason users to assess the expected performance of the platform. However, the orchestrator can be also used to test your own engine.

The repository includes also a tool to run the complete pipeline. Pandoc and textlive packages are required. To do that, simply use these commands:

apt-get update
apt-get install -y pandoc texlive
pip install gnss-benchmark
gnss_benchmark make_report

This tool will take a while to process (ca. 5 minutes), and after that a PDF report should have been generated.

Note that if you do not have LaTEX installed in your system you may not be able to generate the report in PDF. In this case, you can try other formats such as LibreOffice ODT. To do that type

gnss_benchmark make_report --filename report.odt

Use the help of the tool to get more information

gnss_benchmark -h

Running and developing in a container

To make sure you have all necessary components in the system, you can work using Docker containers (recommended).

The first step would be to build the container

docker build -t gnss-benchmark .

# Usage with docker run
docker run -v `pwd`:/gnss_benchmark -w /gnss_benchmark -ti gnss-benchmark bash

# Development with Jupyter
docker run --env-file .env -v `pwd`:/data -w /data -p 8888:8888 -ti gnss-benchmark jupyter notebook

# Usage with docker compose (recommended)
docker-compose -f docker-compose.yml build
docker-compose -f docker-compose.yml run gnss_benchmark bash
python setup.py install
gnss_benchmark make_report -l DEBUG

When using docker-compose remember to place your Jason credentials in an .env file with these contents:

JASON_API_KEY=<api key>
JASON_SECRET_TOKEN=<your secret token>

Although probably not used for the end user, just for debugging purposes, in the event that you have a local instance of the Jason service running in your facilities, you can set the Jason entry point by defining the JASON_API_URL environment variable (along with its corresponding credentials). As an example:

JASON_API_URL=http://192.168.1.54:10000/api
JASON_API_KEY=<api key>
JASON_SECRET_TOKEN=<your secret token>

Using custom processing engine

By default, the tool comes bundled with Rokubun Jason's processing engine, but the user can specify its own processing engine. This cannot be done with the command line tool: a custom Python code will have to be made using the gnss_benchmark module. In this module, there is a package named jason with an example of processing_engine that the user can follow to define other processing engines.