TSNConf

Try the online DEMO.

Feel free to contact me with issues & feedback.

Acknowledgement

This application was developed at the Technical University of Denmark as part of my PhD project. Many thanks to my supervisor Paul Pop and Silviu S. Craciunas for their feedback and support.

Requirements

Python 3.9
GraphViz
- Add to PATH
On Windows:
- Build Tools for Visual Studio: https://visualstudio.microsoft.com/downloads/ (required for some libraries)
  - Only default C++ build tools necessary

Optional requirements

GGen (If you want to generate your own testcases)

Setup

It is recommended to setup a python virtual environment before installing the requirements.

Setup:

git clone https://github.com/nreusch/TSNConf.git
cd TSNConf
pip install -r requirements.txt

Running

Run a simple example:

python main.py .\testcases\TC0_example.flex_network_description --mode 1 --visualize

Then open http://127.0.0.1:8050 in your browser.

The output of each run is stored in testcases\output\XXX\, where XXX is the name of the testcase with a suffix for the chosen parameters. The output includes:

Graphs of all applications and the topology as .gv (GraphViz), .svg & .pdf
A .flex_network_description including the testcase + whatever was generated/optimized(security apps, streams, routing, schedule)
A .pickle of the solution. This can be used to show the dashboard of the results again without rerunning the optimization
If a schedule exists: A schedule.html, the interactive schedule

To show the dashboard for a given .pickle, run this in the root folder:

python -m dashboard.dash_outputter [PICKLE_PATH]

For example:

python -m dashboard.dash_outputter testcases\output\TC0_example_mode1_600_1800_1800_0.2\solution.pickle

Then open http://127.0.0.1:8050 in your browser.

Running extensibility & task mapping optimization

To run a testcase where you want to optimize the extensibility, i.e. have regular free intervals between tasks, and where you have tasks without mapping, use:

-- mode CP_ROUTING_CP_SCHEDULING_EXT

A task without mapping does not have a node assigned in the .flex_network_description. It may have a list of allowed_assignments, which constrains on which this task may be mapped. Make sure that tasks that send & receive streams are only allowed to be mapped to connected ES.

Example task without mapping:

<task name="AppNoMap_t1" wcet="500" period="1000" type="NORMAL" allowed_assignments="ES1,ES2,ES3,ES4"/>

Available commands:

usage: main.py [-h] [--mode MODE] [--aggregate AGGREGATE] [--extra_apps_path EXTRA_APPS_PATH] [--visualize] [--port PORT] [--no_redundancy] [--no_security] [--allow_overlap] [--allow_infeasible] [--original_tesla] [--timeout_routing TIMEOUT_ROUTING]
               [--timeout_scheduling TIMEOUT_SCHEDULING] [--timeout_pint TIMEOUT_PINT] [--k K] [--a A] [--b B] [--Tstart TSTART] [--alpha ALPHA] [--Prmv PRMV] [--w W]
               testcase_path

test

positional arguments:
  testcase_path         The path to the testcase. E.g "testcases/TC0_example.flex_network_description"

optional arguments:
  -h, --help            show this help message and exit
  --mode MODE           Mode to run
  --aggregate AGGREGATE
                        Aggregate results to [path]
  --extra_apps_path EXTRA_APPS_PATH
                        Extra applications to apply onto initial schedule
  --visualize           Visualize testcase (show dashboard)  (Default: False)
  --port PORT           Visualization Dashboard Port
  --no_redundancy       Disable redundancy, by ignoring RL specified in testcase  (Default: False)
  --no_security         Disable security  (Default: False)
  --allow_overlap       Allow redundant stream overlap for CP solutions  (Default: False)
  --allow_infeasible    Allow the CP solver to produce infeasible solutions, including stream overlap or missed deadlines
  --original_tesla      Use TESLA without modification (key only sent in secure stream). Will incure higher bandwidth and cost
  --timeout_routing TIMEOUT_ROUTING
                        Timeout: Routing Solver (s) (Default: 300)
  --timeout_scheduling TIMEOUT_SCHEDULING
                        Timeout: Scheduling Solver (s) (Default: 300)
  --timeout_pint TIMEOUT_PINT
                        Timeout: Pint Solver (s) (Default: 60)
  --k K                 K: Parameter for k-shortest-paths heuristic (Default: 3)
  --a A                 a: Weight for routing overlap component of cost function (Default: 5000)
  --b B                 b: Weight for infeasible stream component of cost function (Default: 1000)
  --Tstart TSTART       Tstart: Start temperature for SA metaheuristic (Default: 10)
  --alpha ALPHA         alpha: Cooling factor for SA  metaheuristic (Default: 0.999)
  --Prmv PRMV           Prmv: Probability of a routing move for SA metaheuristic (Default: 0.8)
  --w W                 w: Weight for punishing overlap in k-shortest-paths for SA heuristic (Default: 10)

Available Modes:
- Mode 0: View/Check the testcase. Doesn't generate/optimize anything
- Mode 1: CP Routing, CP Scheduling, Security, Redundancy, Optimization (Optimize laxity)
- Mode 2: CP Routing, CP Scheduling, Security, Redundancy, Optimization (Optimize laxity + extensibility)
- Mode 11: SA Routing, ASAP Scheduling, Security, Redundancy, Optimization
- Mode 12: SA Routing, SA Scheduling, Security, Redundancy, Optimization
- Mode 13: SA Routing+Scheduling, Security, Redundancy, Optimization
- Mode 99: Check Edge applicaitons against existing solution (Provide solution.pickle as path and extra_apps using --extra_apps_path

nreusch / TSNConf

readme