SimPy Cyber RL Env Test

Classes and Functions

SimpyCyberEnvTest.create_env_run(graph, channelModel, envDebug, queueLimit, channelBW, withThreat)

Initialize the Simpy based cyber environment and run with a random agent..

Parameters

  • graph (graph) – Networkx graph model

  • channelModel (bool) – select if there is channel considered within state space

  • envDebug (bool) – enable/disable the debugging of the cyber environment

  • queueLimit (int) – defined the queue limit of the routers in the network

  • channelBW (int) – set the channel bandwidth

  • withThreat – indicates if threat model is enabled

Returns

env

Return type

gym.Env

SimpyCyberEnvTest.testAgent(env, model)

Test the trained model..

Parameters

  • env (gym.Env) – The openDss RL environment

  • model (gym.Env) – The trained policy network

Returns

Nothing

Return type

None

SimpyCyberEnvTest.trainAgent(env)

Train the model using either A2C or PPO..

Parameters

env (gym.Env) – The openDss RL environment

Returns

The trained model

Return type

Torch model