CCSDS Navigation Data Messages (NDM) Read/Write

Description

The Consultative Committee for Space Data Systems (CCSDS) develops communications and data systems standards for spaceflight. CCSDS Navigation Data Messages (NDM) is the set of file standards to define common data types such as trajectory, orbit, attitude and conjunction events. These data types are routinely generated and exchanged within and between spacecraft operators, space agencies, researchers, amateurs and commercial companies. As such, accurate definition and common interpretation of the data is crucial (and sometimes mission-critical).

The standard description for each data type is encapsulated in an XML Schema file. This project ccsds-ndm aims to be the reference open-source Python implementation to read and write the NDM XML and KVN files, through an object tree API, auto-generated by these schema files. It supports the up-to-date NDM XML 4.0 standard as well as the old NDM XML 2.0 and 1.0 standards.

The source code is on Github.

Current functionality:

	Read	Write
XML	All NDM Types	All NDM Types
KVN	All except NDM Combined Instantiation	All except NDM Combined Instantiation
JSON	Not specified in CCSDS Standards	Not specified in CCSDS Standards

Usage and Examples

There are two main use cases:

The ccsds-ndm library reads the NDM file, fills an object tree and offers it to the users. The users will then have to fill their own attitude, orbit or trajectory objects used in their libraries.
The user fills an object tree from their own attitude, orbit or trajectory object. The ccsds-ndm library writes the NDM file using this object tree.

For the first use case, reading an OEM file from xml_read_path is as simple as:

cdm = NdmIo().from_path(xml_read_path)

Note that file format (XML or KVN) and data type (e.g. CDM or NDM) are inferred automatically. The output cdm is the object tree for a Conjunction Data Message (CDM). The contents can then be reached going deeper in the object tree as specified in the corresponding NDM Standard. This example shows how to reach the orbit normal position component of the relative state vector:

print(cdm.body.relative_metadata_data.relative_state_vector.relative_position_n)

The data can sometimes be of type NDM Combined Instantiation. This means that there are multiple NDM data bodies (e.g. 2 AEMs, 3 OEMs and one OMM) within a single file. The file reader supports these types as well and they are kept within the Ndm object as individual lists for each of the file types. The following example retrieves the second Omm object in the NDM file and then continues to dive deeper into the object tree to retrieve the eccentricity value.

print(ndm.omm[1].body.segment.data.mean_elements.eccentricity)

If the file is of the type NDM Combined Instantiation but there is only a single data (e.g. OMM) in it, the ndm tags are stripped and only the single data is presented to the user.

Filling the objects with data properly requires some care. As the standard is understandably strict, the object tree derived from the XSD files are also rather exacting in how they accept data. Every field that holds a physical quantity must be set with the corresponding wrapper type (e.g. LengthType, DvType). Assigning a plain number or string raises a TypeError immediately.

# Correct: wrapper type with explicit units
cdm.body.relative_metadata_data.relative_state_vector.relative_position_t = LengthType(value=800, units=LengthUnits.M)

# Also correct: pint or astropy Quantity — auto-wrapped to the right type
import pint
u = pint.UnitRegistry()
cdm.body.relative_metadata_data.relative_state_vector.relative_position_r = 700 * u.m

# Also correct with auto-convert enabled: any compatible unit is silently converted
from ccsds_ndm.model_quantity import set_auto_convert
set_auto_convert(True)
cdm.body.relative_metadata_data.relative_state_vector.relative_position_r = 0.7 * u.km  # → 700.0 m

# Invalid: raises TypeError
cdm.body.relative_metadata_data.relative_state_vector.relative_position_r = 600

The XML output for the correct assignments properly transmits unit information:

<relativeStateVector>
  <RELATIVE_POSITION_R units="m">700</RELATIVE_POSITION_R>
  <RELATIVE_POSITION_T units="m">800</RELATIVE_POSITION_T>

Optional Unit Support: Reading the value back as a pint or astropy Quantity is done via the .q() method on any wrapper instance:

from ccsds_ndm.model_quantity import set_quantity_mode, QuantityMode
set_quantity_mode(QuantityMode.PINT)

q = cdm.body.relative_metadata_data.relative_state_vector.relative_position_t.q()
# q is a pint Quantity: 800.0 meter

Therefore, care must be taken (and standard documents must be kept as a reference) when mapping the user objects into the object tree. Valuable information on units, models and methods can be found there to correctly interpret the data. Also, comments can also be used to provide supplementary information on how this data is generated. Full details on validation and quantity support are in Type Validation and Quantity Support.

Finally, once filled with the relevant data, the cdm object can be written to xml_write_path in XML data format as:

NdmIo().to_file(cdm, NDMFileFormats.XML, xml_write_path)

The ndm object trees are not very user friendly and most probably will have to be filled by the users’ own equivalent objects (trajectory, orbit, attitude etc.).

Installing `ccsds-ndm`

The ccsds-ndm package is on PyPI and you can install it simply by running:

pip install ccsds_ndm

You can also install it via conda-forge:

conda install -c conda-forge ccsds_ndm

Do not install ccsds-ndm using sudo.

Requirements

xsData is used to read and write XML files (and also to generate the object tree)
lxml to support XML object creation
pint (optional) for pint Quantity input/output — install with pip install ccsds_ndm[pint]
astropy (optional) for astropy Quantity input/output — install with pip install ccsds_ndm[astropy]

Citation

Please use the DOI for citations. This is the latest version:

Known Issues

Some browsers (at least Firefox 82.0 in Dec 2020) strip the namespace information from the XML file when the file is viewed. However, when the file is downloaded directly (e.g. through wget), the namespace information is kept intact. While neither case causes issues with ccsds-ndm, the former case may result in the XML viewers to report namespace errors. (Hat tip to Juan Luis Cano Rodríguez for spotting this)