The use of NGSO satellite constellations offers many advantages for the communication missions, in particular potential global coverage (possibly adapted to demand density), low delay transmission, robustness of the system, and potential low cost of gigabit per second.
To meet non uniform and evolving user demand as well as coordination with other systems and regulatory constraints, payload flexibility and reconfigurability will be key features of on-going and future constellation projects.
Thus, besides the constellation dynamics inducing constant changes in the user or feeder links properties and frequent handovers, the system flexibility features and constraints need also to be considered for the radio-resource management which lead to a highly complex problem to optimize.
In this study, the objective is then to take into account all these aspects to design, implement and evaluate an RRM algorithm that shall provide the highest possible performance (in regards with defined KPIs) whilst dealing with operational constraints (particularly in terms of computation time).
The first challenge of the project was the choice of the algorithms to develop, given the constraints of time calculations, and the objective of maximizing the constellation throughput.
The second challenge was the optimization of the simulator speed of calculation of the interferences.
The third challenge was the choice and the design of the system scenarios, to be as close as possible to a real one
The simulator is able to emulate a large non-geostationary constellation and to evaluate the resource allocation algorithm by computing the total throughput considering the interferences. The simulator is highly modular, it can use any antenna pattern, satellite and user position, etc.. It is made of multiple modules, that could be replaced if needed later.
The resource allocation algorithm is able to handle a large quantity of users and satellites, and to allocate resources with good overall performances.
The product is composed of the following components: a simulator to emulate constellations and algorithms to maximise its throughput.
The product works as follow:
The simulator initialize itself by reading input files and creating the constellation.
Then the simulator enters a loop, where for each time step of the simulation it calls the RRM algorithm, which allocates each cluster
It computes the actual throughput for each cluster (including the interferences computation), and the total throughput of the constellation, as well as a lot of other statistics and information about the constellation state.
The simulator is designed to be modular, every component and algorithm can be upgraded later on
This project emulates a non-geo stationary constellation with a powerful simulator. This simulator is based on multiple modules, where each module either reads inputs from simulation tools, or computes precise information, such as interferences. Finally, the simulator has the capability to export multiple outputs (statistics, figures …). That can be configured by the user through configuration files.
The development logic follows these steps:
System scenario definition
Technical requirement specification
RRM module design
RRM demonstrator design
RRM demonstrator development
RRM performance assessment
Regarding the industrial organisation, Magister is responsible of the design and development of the simulator, while Thales Alenia Space is responsible of the overall project, including the definition of the system and of the algorithm of the RRM module.
The project has been completed.