With small satellites moving past LEO and targeting interplanetary missions, a whole set of new components is becoming part of the standard platform hardware: TT&C transponders. Traditionally small LEO spacecraft rely on different means to estimate their position (like TLEs or GPS measurements) but these mechanisms are not available in deep space, where satellites must rely on the traditional radiometric tracking. Transponders implementing ranging function and coherent downlink are required to best estimate the spacecraft position and relative velocity. This represents an evolution of the standard transceivers used in CubeSats that so far had limited need for tracking capabilities as navigation was not concerned or achieved with other systems.
In this paper we focus on the efforts to develop a fully software-defined transponder compatible with the ESTRACK network to enable a fully European interplanetary CubeSat mission. We will present the lessons learnt on using tools like GNU Radio and other open-source applications for space applications, dealing with the typical testing, qualification and quality assurance issues.
Due to the complexity of the testing and qualification of TT&C transponders, a fully software-defined testbed has been designed, capable of emulating the complete radio chain starting from the ground modem to the interplanetary RF channel till the satellite transponder. This solution allows to quickly develop the system and evaluate its performances before starting the real hardware development.
Using existing SDR platforms, real hardware can be easily added in the loop. This approach allows to use actual units (existing ground modems or satellite transponders, for example) in the simulation testbed to speed-up design and characterize performances. Such a testbed can perform automated tests on the complete system, leading to a very quick unit acceptance process, as needed especially for manufacturing transponders on a large scale.
The testbed has been designed using GNU Radio, as it allows to reuse a large library of existing functions, but it was targeted towards a user base accustomed with space systems development rather than deep SDR algorithmic knowledge. This has been done developing two modules implementing ECSS-specific functions and relevant standard terminology. Dedicated wrapper components re-using existing GNU Radio functionalities with customized interfaces or new blocks for missing functionalities have been developed. One of the missing functionalities was, for example, the carrier recovery block capable to generate the downlink carrier coherent with uplink one but on a different frequency with a fixed and accurate ratio. GNU Radio for simplicity uses 32-bit floating point math in existing blocks and this might limit the actual accuracy of the transponder turn-around ratio. To prevent this, integer math has been implemented to mimic the behaviour of real hardware. GNU Radio is also focused mainly on existing communication standards so typical components and algorithms used in the space domain are not present and were implemented (for example Data Transition Tracking Loop for clock recovery). This allows to verify the performances of standard algorithms and compare it with more advanced ones (already present in GNU Radio) but of not common use in space applications. A dedicated component capable of performing generic orbital simulations (based on existing open-source solutions) has been added and connected to GNU Radio.
An important part of the effort has been spent focusing on the quality assurance issues arising from a public open-source project with multiple developers and a very fast update pace to be used in an aerospace environment. A complete set of unit and integration tests has been developed to ensure also core functionalities are validated against known data sets to prevent bugs introduced in future versions. An automated test reporting system has also been developed to prevent manual errors from entering the quality assurance chain, and a full file checksum system has been created to guarantee integrity of the files with respect to the qualified version.

ITP Office is the Radiofrequency Instrumentations, Software Defined Radio, Telemetry & Telecommand and Propagation Office within the RadioFrequency Department (RF/ITP) in CNES.

ITP Office is responsible for the RF chains design for CNES satellites and in particular of the specification of the satellite S-Band and X-Band chains in order to respect the following CCSDS requirement with the multimission CNES ground stations network:
The link budget shall be established with an elevation angle of 5° and shall identify the following margins
- for TM and TC in nominal mode: 3 dB minimum in nominal case, 0 dB in worst case.
- for TM and TC in safe mode: 3 dB minimum in nominal case, 0 dB in worst case.
Note: “nominal case” takes into account nominal parameters of link budget; “worst case” takes into account worst case values (EOL, temperature, altitude…) pondered versus RSS method.

Our traditional tool for link budget was based on the CCSDS requirements and was complying with RSS method.
But in some cases, in particular for survival case, the computed margins appear very weak or even negative. We analyzed the CCSDS method and its representability in order to understand these weak margins. We found out that the worst case calculated with the CCSDS method was very pessimistic and did not reflect the statistical reality. In other words, the worst link budget could happen, but not with the same occurrence as calculated with the RSS method. This comes from the statistical repartition of several parameters. For instance, the satellite antenna gains: the statistical repartition proposed by the CCSDS method (triangular) is not representative of the reality in survival mode. So the link budget was computed with very pessimistic values and the margins were abnormally low.

MICROCARB is a CNES LEO mission. Its purpose is to map sources and sink of CO2 on a global scale. The satellite Equipment include a S-Band transponder and a HDRT X-Band Transmitter”.
Microcarb project asked also to the ITP Office to compute the availability of the end to end transmission. The availability depends on the method used for the link budget computation, but also on the availability hypothesis related to propagation losses (rains, absorption of oxygen etc..). The link budget (with the traditional tool) used to be computed at 3 sigma, taking into account a propagation availability of 99%. For the record, the CNES DLL propagation tool allows calculating propagation losses from a fixed availability, knowing the satellite elevation, the signal frequency and the ground station geographical characteristics and location. A 3-sigma link budget with a propagation availability of 99% corresponds to a global availability of 98.8% which is not what was taken into account in the global satellite availability calculation. Thus, the disadvantage of the traditional method comes from the choice of the propagation availability, a priori

In order to compute this availability with a realistic link budget, a new tool was developed, based on statistical random draw. For each link budget case, the tool computes one million link budgets, picking the value of each parameters among the real values and taking into account the statistical repartition of this value. For example: the value of antenna gain in survival mode, is picked in all the coverage. Whereas, in nadir mode, the antenna gain value is picked among all the values corresponding to the area of the coverage related to the satellite elevation.

For availability computation with the new tool, the idea is to use the margin of the link budget to deduce the propagation availability that can be reached:
- The million link budgets (done with no propagation losses) allow obtaining a probability distribution for the Eb/N0: “Eb/N0_rule”.
- For an elevation “El” of the satellite and a given ground station, thanks to the CNES DLL propagation tool the probability distribution of the propagation losses can be computed: called “Propagation_rule”.
- To each value of Eb/N0 corresponds a value refered as “margin” (difference with the required Eb/N0) which can be used to determine the “ponderation factor” = the probability of “margin” in “Propagation_Rule”.
Then, the “Eb/N0_rule”, ponderated by “ponderation factor” for each value of Eb/N0, provides a probability distribution of Eb/N0 that takes into account propagation losses. The global availability is calculated thanks to these values.

This new tool allows obtaining a realistic link budget and computing the corresponding availability. It was used for Microcarb link budgets, including HDRT in X-Band.

The article will be organized in three chapters:
In a first part, we will explain the CCSDS method for link budget, for TT&C and HDRT, and then we will analyze this method.
In a second part, we will present our new tool of statistical method for link budget and availability computation.
At last, we will give an example for Microcarb TT&C link budget in survival and nadir modes.