Some path schedule solutions specifically for UAV-sensorTime-sensitive Network. The code is used to evaluate the models' performance and time complexity with three different algorithms. Especially, establishing the DQN solving.
- length_range: The range of map's length and width.
- priority_range: The range of possible sensors' priority.
- s:
- v: The speed of UAV.
- period: One period named
t
in the paper. - t_limit: The working-time limitation which is caused by fuel.
- max_time: The total time the system run.
- seed: Random number seed for generate random sensors lists.
-
Install Python and dependencies with Miniconda
$ conda -V conda 4.7.10
-
Install Python and dependencies with conda.
$ conda env create -f environment.yml $ conda activate UAV (UAV) $ python -V Python 3.7.3
or
$ conda env update -f environment.yml
-
Run Python script. The results save at
./out/
(UAV) $ python greedy.py (UAV) $ python QL.py (UAV) $ python DQN.py
-
If you wanna remove conda environment:
(UAV) $ conda deactivate $ conda remove -n UAV --all