minspp

Minimalistic implementation of the Space Packet specification from the CCSDS Space Packet Protocol standard.

Repository | Documentation

This package was formerly published as minsp, which is no longer updated.

Installation

Install using pip:

$ pip install minspp

Install package from the git repository:

$ pip install git+https://github.com/nunorc/minspp@master

Getting Started

Import the SpacePacket class from the package:

>>> from minspp import SpacePacket

For example, to create a new space packet for APID 11 and an arbitrary data field:

>>> space_packet = SpacePacket(apid=11, data_field=b'hello')
>>> space_packet
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=0, apid=11, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=4, secondary_header=b'', data_field=b'hello')

To get the bytes representation of the packet:

>>> byte_stream = space_packet.as_bytes()
>>> byte_stream
b'\x00\x0b\xc0\x00\x00\x04hello'

Packets can also be created from a byte stream:

>>> new_packet = SpacePacket.from_bytes(byte_stream)
>>> new_packet
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=0, apid=11, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=4, secondary_header=b'', data_field=b'hello')
>>> new_packet.data_field
b'hello'

The packet data length field delimits the packet, so any octets past the end of the packet are ignored. To walk a buffer holding several back to back packets use iter_packets, which takes the same secondary header arguments as from_bytes:

>>> stream = b'\x00\x0b\xc0\x00\x00\x04hello\x00\x0c\xc0\x00\x00\x04world'
>>> [(p.apid, p.data_field) for p in SpacePacket.iter_packets(stream)]
[(11, b'hello'), (12, b'world')]

The packet data field can end with a packet error control field, the CRC-16-CCITT of every preceding octet of the packet, which most missions mandate for telecommands. It is opt-in on both sides, and the data length written to the primary header accounts for the two extra octets:

>>> byte_stream = space_packet.as_bytes(packet_error_control=True)
>>> byte_stream
b'\x00\x0b\xc0\x00\x00\x06hello\x81c'
>>> SpacePacket.from_bytes(byte_stream, packet_error_control=True).data_field
b'hello'

Decoding verifies the field and strips it, a mismatch raises a ValueError.

Secondary header can have a custom data definition, or to use PUS. Telemetry packets use the PUS-C (ECSS-E-ST-70-41C) TM secondary header:

>>> from minspp.pus import PUSTMHeader
>>> pus_header = PUSTMHeader()
>>> pus_header
PUSTMHeader(version=2, time_reference_status=0, service_type=1, service_subtype=1, message_type_counter=0, destination_id=0, has_time=False, cuc_time=b'')

And create a new packet with the PUS header:

>>> space_packet = SpacePacket(secondary_header=pus_header)
>>> space_packet
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=1, apid=0, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=6, secondary_header=PUSTMHeader(version=2, time_reference_status=0, service_type=1, service_subtype=1, message_type_counter=0, destination_id=0, has_time=False, cuc_time=b''), data_field=b'')

For example a housekeeping parameter report (service 3, subtype 25) for destination 42:

>>> tm_header = PUSTMHeader(service_type=3, service_subtype=25, message_type_counter=7, destination_id=42)
>>> tm_header.as_bytes()
b' \x03\x19\x00\x07\x00*'

The width of the TM destination ID is mission defined, use destination_id_length to set it in bytes (defaults to 2, and 0 means the field is absent):

>>> PUSTMHeader(service_type=3, service_subtype=25, message_type_counter=7, destination_id=42, destination_id_length=1).as_bytes()
b' \x03\x19\x00\x07*'

The split of the CUC time between the coarse (seconds) and the fine (sub-seconds) field is mission defined too, use cuc_coarse_length to set the coarse part, the remaining octets of cuc_time_length hold the fine part:

>>> header = PUSTMHeader(has_time=True, cuc_time_length=7, cuc_coarse_length=5)
>>> header.fine_time_length()
2

So is the epoch the coarse time counts from, which defaults to the Unix epoch, use cuc_epoch for a mission that counts from another one (the CCSDS agency-standard epoch is 1958-01-01, and mission elapsed time counts from launch):

>>> from datetime import datetime, timezone
>>> from minspp.utils import cuc_as_datetime
>>> epoch = datetime(1958, 1, 1, tzinfo=timezone.utc)
>>> header = PUSTMHeader(has_time=True, cuc_epoch=epoch)
>>> cuc_as_datetime(header.cuc_time, epoch=epoch)   # the current time
datetime.datetime(2026, 7, 29, 20, 17, 2, 120716, tzinfo=datetime.timezone.utc)

The CUC field carries neither the epoch nor the coarse and fine lengths, they are declared out-of-band, so decoding must use the same values the sender used. There is no P-field (preamble) support, and all the arithmetic is done on UTC datetimes: a mission whose CUC counts a continuous time scale (TAI, GPS) is off by the leap seconds accumulated since its epoch, this package does not convert time scales.

Telecommand packets use a different secondary header layout, implemented by the PUSTCHeader class:

>>> from minspp.pus import PUSTCHeader
>>> PUSTCHeader()
PUSTCHeader(version=2, ack=0, service_type=1, service_subtype=1, source_id=0, has_time=False, cuc_time=b'')

The width of the TC source ID is mission defined, use source_id_length to set it in bytes (defaults to 1, and 0 means the field is absent):

>>> PUSTCHeader(service_type=8, service_subtype=1, source_id=0x0102, source_id_length=2).as_bytes()
b' \x08\x01\x01\x02'

Field values are checked when a packet or a header is packed, not when it is built. An out of range value raises a ValueError instead of being silently masked or surfacing as a struct.error, and the standard reserves service type and message subtype 0:

>>> PUSTCHeader(service_type=300).as_bytes()
Traceback (most recent call last):
  ...
ValueError: Invalid service type 300, must be between 1 and 255.

Decoding stays permissive, so a malformed packet can still be inspected. Use strict (or pus_strict on SpacePacket.from_bytes) to reject a secondary header that is not valid PUS-C, i.e. one whose version is not 2 or whose service type or subtype is the reserved 0:

>>> PUSTCHeader.from_bytes(b'\x10\x08\x01\x00').version   # a PUS-A header
1
>>> PUSTCHeader.from_bytes(b'\x10\x08\x01\x00', strict=True)
Traceback (most recent call last):
  ...
ValueError: Invalid PUS version 1, must be 2 for PUS-C.

Note that ECSS-E-ST-70-41C defines no time field for the TC secondary header, timestamps belong to telemetry. The has_time and cuc_time arguments of PUSTCHeader are a mission specific extension for missions that extend the header, they are off by default and leaving them off keeps the header standard conformant.

Similar approach for a MAL secondary header:

>>> from minspp.mo import MALHeader
>>> mal_header = MALHeader()
>>> mal_header
MALHeader(version=0, sdu_type=0, service_area=0, service=0, operation=0, area_version=0, is_error=0, qos_level=0, session=0, secondary_apid=0, secondary_apid_qualifier=0, transaction_id=0, source_id_flag=0, destination_id_flag=0, priority_flag=0, timestamp_flag=0, network_zone_flag=0, session_name_flag=0, domain_flag=0, authentication_id_flag=0, source_id=0, destination_id=0, segment_counter=0, priority=0, timestamp=None, network_zone='', session_name='', domain='', authentication_id='')

And to create a new packet with the MAL header:

>>> space_packet = SpacePacket(secondary_header=mal_header)
>>> space_packet
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=1, apid=0, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=20, secondary_header=MALHeader(version=0, sdu_type=0, service_area=0, service=0, operation=0, area_version=0, is_error=0, qos_level=0, session=0, secondary_apid=0, secondary_apid_qualifier=0, transaction_id=0, source_id_flag=0, destination_id_flag=0, priority_flag=0, timestamp_flag=0, network_zone_flag=0, session_name_flag=0, domain_flag=0, authentication_id_flag=0, source_id=0, destination_id=0, segment_counter=0, priority=0, timestamp=None, network_zone='', session_name='', domain='', authentication_id=''), data_field=b'')

To create a space packet from a byte stream including a PUS header, use pus_tm=True for a TM header, or pus_tc=True for a TC one:

>>> byte_stream = SpacePacket(secondary_header=pus_header).as_bytes()
>>> SpacePacket.from_bytes(byte_stream, pus_tm=True)
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=1, apid=0, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=6, secondary_header=PUSTMHeader(version=2, time_reference_status=0, service_type=1, service_subtype=1, message_type_counter=0, destination_id=0, has_time=False, cuc_time=b''), data_field=b'')

A TC header with a non default source ID width must be decoded with the same width, otherwise the data field is misaligned, use pus_source_id_length:

>>> tc_header = PUSTCHeader(service_type=8, service_subtype=1, source_id=0x0102, source_id_length=2)
>>> byte_stream = SpacePacket(secondary_header=tc_header, data_field=b'\xAB\x2A').as_bytes()
>>> SpacePacket.from_bytes(byte_stream, pus_tc=True, pus_source_id_length=2).data_field
b'\xab*'

The same holds for the other mission defined values, use pus_destination_id_length for a TM header, pus_cuc_coarse_length for the coarse and fine time split, and pus_cuc_epoch for the epoch:

>>> tm_header = PUSTMHeader(service_type=3, service_subtype=25, destination_id=42, destination_id_length=1)
>>> byte_stream = SpacePacket(secondary_header=tm_header, data_field=b'\x01\x02').as_bytes()
>>> SpacePacket.from_bytes(byte_stream, pus_tm=True, pus_destination_id_length=1).data_field
b'\x01\x02'

Or from a byte stream including a MAL header:

>>> byte_stream = SpacePacket(secondary_header=mal_header).as_bytes()
>>> SpacePacket.from_bytes(byte_stream, mal=True)
SpacePacket(version=0, type=<PacketType.TM: 0>, secondary_header_flag=1, apid=0, sequence_flags=<SequenceFlags.UNSEGMENTED: 3>, sequence_count=0, data_length=20, secondary_header=MALHeader(version=0, sdu_type=0, service_area=0, service=0, operation=0, area_version=0, is_error=0, qos_level=0, session=0, secondary_apid=0, secondary_apid_qualifier=0, transaction_id=0, source_id_flag=0, destination_id_flag=0, priority_flag=0, timestamp_flag=0, network_zone_flag=0, session_name_flag=0, domain_flag=0, authentication_id_flag=0, source_id=0, destination_id=0, segment_counter=0, priority=0, timestamp=None, network_zone='', session_name='', domain='', authentication_id=''), data_field=b'')

Use SpacePacketAssembler to recover the data from a list of fragmented packets, for example consider the following packets:

>>> from minspp import SpacePacket, SequenceFlags
>>> sp1 = SpacePacket(sequence_flags=SequenceFlags.FIRST, data_field=b"123")
>>> sp2 = SpacePacket(sequence_flags=SequenceFlags.CONTINUATION, data_field=b"456")
>>> sp3 = SpacePacket(sequence_flags=SequenceFlags.LAST, data_field=b"789")

To recover the fragmented payload by processing the individual packets:

>>> from minspp import SpacePacketAssembler
>>> spa = SpacePacketAssembler()
>>> spa.process_packet(sp1)
>>> spa.process_packet(sp2)
>>> spa.process_packet(sp3)
b'123456789'

Or directly using the from_packets method:

>>> SpacePacketAssembler.from_packets([sp1, sp2, sp3])
b'123456789'

Acknowledgements

• Dominik Marszk for general support and MAL header baseline implementation.