A modular DTN simulator in the style of the ONE.
Features:
- distance- or contact-window-based network simulation
- DTN routing algorithms
- epidemic
- spray & wait
- first contact
- direct delivery
- PRoPHET
- static
- mobility
- random waypoint
- external ONE movement
- external ns2 movement
- contact plan connectivity model
- ION DTN contact plans
- [core contact plan}(https://github.com/gh0st42/ccm/)
- static networkx topology
- optionally: from graphml
- optionally: fluctuating from contact plan
- simulated user applications
- tools
neteditfor generating graphml topologiesponsanimfor generating animated gifs and mp4 from graphml topologies with a contact plan or event logsscenariorunnerto simulate CCSDS DTN reference scenarios without writing any code
The scenariorunner consumes the scenario format of the CCSDS DTN Reference
Scenarios (see the vendored copies and usage instructions in
examples/scenarios/ccsds):
nodes.json— nodes withnode_label,node_nameandipn:node_idplanned_contacts.csv— contact plan used as routing knowledgeactual_contacts.csv— optional data-plane contacts (defaults to the planned contacts)traffic.json— application traffic flows (src,dst,adu_size,interval, optionalttl)
Run a scenario directory directly:
scenariorunner --scenario-dir examples/scenarios/ccsds/lunar-communication/v1.0 -r StaticRouter -t 86400or specify the files individually with --nodes, --contacts
, --actual-contacts and --flows.
Run the test suite (unit tests for the parsers and scenario loaders as well as all example scripts):
pip install -e ".[test]"
python -m pytestCI runs the same suite via GitHub Actions on every push and pull
request (see .github/workflows/tests.yml).
- simpy >= 4.0
- networkx >= 3.2
- plotting:
- seaborn
- pandas
- matplotlib
- numpy
- tools:
- pillow
- opencv-python
- tkinter
import random
import json
import pons
import pons.routing
RANDOM_SEED = 42
SIM_TIME = 3600*24
NET_RANGE = 50
NUM_NODES = 10
WORLD_SIZE = (3000, 3000)
# Setup and start the simulation
random.seed(RANDOM_SEED)
moves = pons.generate_randomwaypoint_movement(
SIM_TIME, NUM_NODES, WORLD_SIZE[0], WORLD_SIZE[1], max_pause=60.0)
net = pons.NetworkSettings("NET1", range=NET_RANGE)
epidemic = pons.routing.EpidemicRouter()
nodes = pons.generate_nodes(NUM_NODES, net=[net], router=epidemic)
config = {"movement_logger": False, "peers_logger": False, "event_logger": True}
msggenconfig = {"type": "single", "interval": 30,
"src": (0, NUM_NODES), "dst": (0, NUM_NODES),
"size": 100, "id": "M"}
netsim = pons.NetSim(SIM_TIME, WORLD_SIZE, nodes, moves,
config=config, msggens=[msggenconfig])
netsim.setup()
netsim.run()
# print results
print(json.dumps(netsim.net_stats, indent=4))
print(json.dumps(netsim.routing_stats, indent=4))Run using python3 or for improved performance use pypy3.
Some of the simulation core functions can be set during runtime without having to change your simulation code.
LOG_FILEcan be set to change the default event log file from/tmp/events.logto something elseSIM_DURATIONcan be used to override the calculated simulation duration
For netedit there are also ways to influence its behavior:
BG_IMGcan be set to any image and it while be rendered as a background behind the network topology