Broadcast Communication

By default, communication occurs between all satellites that are specified by the communication method. This tutorial shows how to use two classes to configure a broadcast action that must be taken for communication to occur:

Broadcast, which gives satellites an action that enables communication from them at the end of the current step.
BroadcastCommunication, which can be combined with another communication method to limit communication from broadcasters to those satellites satisfying the requirements of the other communication method.

Configuring the Environment

For this example, a multisatellite target imaging environment will be used. The goal is to maximize the value of unique images taken. This configuration is similar to the Multi-Agent Environments example.

[1]:

from bsk_rl import sats, act, obs, scene, data, comm
from bsk_rl.sim import dyn, fsw


class ImagingSatellite(sats.ImagingSatellite):
    observation_spec = [
        obs.OpportunityProperties(
            dict(prop="priority"),
            dict(prop="opportunity_open", norm=5700.0),
            n_ahead_observe=4,
        )
    ]
    action_spec = [act.Broadcast(duration=15.0), act.Image(n_ahead_image=4)]
    dyn_type = dyn.FullFeaturedDynModel
    fsw_type = fsw.SteeringImagerFSWModel


ACTION_BROADCAST = 0
ACTION_IMAGE_0 = 1
ACTION_IMAGE_1 = 2
ACTION_IMAGE_2 = 3
ACTION_IMAGE_3 = 4

Satellite properties are set to give the satellite near-unlimited power and storage resources. To randomize some parameters in a correlated manner across satellites, a sat_arg_randomizer is set and passed to the environment. In this case, the satellites are distributed in a trivial single-plane Walker-delta constellation.

[2]:

from bsk_rl.utils.orbital import walker_delta_args

N_AGENTS = 2
sat_args = dict(
    imageAttErrorRequirement=0.01,
    imageRateErrorRequirement=0.01,
    batteryStorageCapacity=1e9,
    storedCharge_Init=1e9,
    dataStorageCapacity=1e12,
    u_max=0.4,
    K1=0.25,
    K3=3.0,
    omega_max=0.087,
    servo_Ki=5.0,
    servo_P=150 / 5,
)
sat_arg_randomizer = walker_delta_args(
    altitude=800.0, inc=60.0, n_planes=1, clustersize=N_AGENTS, clusterspacing=5.0
)

A communication type is defined that uses multidegree line-of-sight communication with broadcasting.

[3]:

class BroadcastLOS(comm.BroadcastCommunication, comm.LOSMultiCommunication):
    pass

Finally, the environment can be instantiated.

[4]:

from bsk_rl import ConstellationTasking

env = ConstellationTasking(
    satellites=[ImagingSatellite(f"EO-{i + 1}", sat_args) for i in range(N_AGENTS)],
    scenario=scene.UniformTargets(1000),
    rewarder=data.UniqueImageReward(),
    communicator=BroadcastLOS(),
    sat_arg_randomizer=sat_arg_randomizer,
    log_level="INFO",
)
_ = env.reset()

2025-08-14 23:28:18,724 gym                            INFO       Resetting environment with seed=1530303493
2025-08-14 23:28:18,726 scene.targets                  INFO       Generating 1000 targets
2025-08-14 23:28:18,863 sats.satellite.EO-1            INFO       <0.00> EO-1: Finding opportunity windows from 0.00 to 600.00 seconds
2025-08-14 23:28:18,902 sats.satellite.EO-2            INFO       <0.00> EO-2: Finding opportunity windows from 0.00 to 600.00 seconds
2025-08-14 23:28:18,958 gym                            INFO       <0.00> Environment reset

In the first step, both agents are tasked with an imaging action, and one successfully images a target.

[5]:

_ = env.step({"EO-1": ACTION_IMAGE_3, "EO-2": ACTION_IMAGE_2})

2025-08-14 23:28:18,963 gym                            INFO       <0.00> === STARTING STEP ===
2025-08-14 23:28:18,964 sats.satellite.EO-1            INFO       <0.00> EO-1: target index 3 tasked
2025-08-14 23:28:18,964 sats.satellite.EO-1            INFO       <0.00> EO-1: Target(tgt-697) tasked for imaging
2025-08-14 23:28:18,966 sats.satellite.EO-1            INFO       <0.00> EO-1: Target(tgt-697) window enabled: 33.7 to 211.2
2025-08-14 23:28:18,966 sats.satellite.EO-1            INFO       <0.00> EO-1: setting timed terminal event at 211.2
2025-08-14 23:28:18,967 sats.satellite.EO-2            INFO       <0.00> EO-2: target index 2 tasked
2025-08-14 23:28:18,968 sats.satellite.EO-2            INFO       <0.00> EO-2: Target(tgt-697) tasked for imaging
2025-08-14 23:28:18,970 sats.satellite.EO-2            INFO       <0.00> EO-2: Target(tgt-697) window enabled: 0.0 to 121.8
2025-08-14 23:28:18,970 sats.satellite.EO-2            INFO       <0.00> EO-2: setting timed terminal event at 121.8
2025-08-14 23:28:18,993 sats.satellite.EO-2            INFO       <51.00> EO-2: imaged Target(tgt-697)
2025-08-14 23:28:18,996 data.base                      INFO       <51.00> Total reward: {'EO-2': 0.3287332165996737}
2025-08-14 23:28:18,997 sats.satellite.EO-2            INFO       <51.00> EO-2: Satellite EO-2 requires retasking
2025-08-14 23:28:19,000 gym                            INFO       <51.00> Step reward: {'EO-2': 0.3287332165996737}

When both select the broadcast action, data is shared in both directions. This is subject to line-of-sight availability as well as selecting the correct action.

[6]:

_ = env.step({"EO-1": ACTION_BROADCAST, "EO-2": ACTION_BROADCAST})

2025-08-14 23:28:19,004 gym                            INFO       <51.00> === STARTING STEP ===
2025-08-14 23:28:19,005 sats.satellite.EO-1            INFO       <51.00> EO-1: setting timed terminal event at 66.0
2025-08-14 23:28:19,006 sats.satellite.EO-2            INFO       <51.00> EO-2: setting timed terminal event at 66.0
2025-08-14 23:28:19,011 sats.satellite.EO-1            INFO       <60.00> EO-1: imaged Target(tgt-697)
2025-08-14 23:28:19,013 data.base                      INFO       <60.00> Total reward: {}
2025-08-14 23:28:19,014 comm.communication             INFO       <60.00> Communicating data in 2 directions.
2025-08-14 23:28:19,015 comm.communication             INFO       <60.00> Optimizing data communication between all pairs of satellites
2025-08-14 23:28:19,017 sats.satellite.EO-1            INFO       <60.00> EO-1: Satellite EO-1 requires retasking
2025-08-14 23:28:19,019 gym                            INFO       <60.00> Step reward: {}

One agent can broadcast while the others are completing a different task.

[7]:

_ = env.step({"EO-1": ACTION_IMAGE_2, "EO-2": ACTION_BROADCAST})  # Agent 1 broadcasts

2025-08-14 23:28:19,024 gym                            INFO       <60.00> === STARTING STEP ===
2025-08-14 23:28:19,024 sats.satellite.EO-1            INFO       <60.00> EO-1: target index 2 tasked
2025-08-14 23:28:19,025 sats.satellite.EO-1            INFO       <60.00> EO-1: Target(tgt-200) tasked for imaging
2025-08-14 23:28:19,026 sats.satellite.EO-1            INFO       <60.00> EO-1: Target(tgt-200) window enabled: 0.0 to 139.9
2025-08-14 23:28:19,027 sats.satellite.EO-1            INFO       <60.00> EO-1: setting timed terminal event at 139.9
2025-08-14 23:28:19,028 sats.satellite.EO-2            INFO       <60.00> EO-2: setting timed terminal event at 75.0
2025-08-14 23:28:19,036 sats.satellite.EO-2            INFO       <75.00> EO-2: timed termination at 75.0 for broadcast
2025-08-14 23:28:19,038 data.base                      INFO       <75.00> Total reward: {}
2025-08-14 23:28:19,039 comm.communication             INFO       <75.00> Communicating data in 1 direction.
2025-08-14 23:28:19,040 sats.satellite.EO-2            INFO       <75.00> EO-2: Satellite EO-2 requires retasking
2025-08-14 23:28:19,042 gym                            INFO       <75.00> Step reward: {}