Resource-efficient space observation
The responsible handling of resources demands a fundamental rethink with regard to the design of future space observation systems. New concepts can now be systematically examined for the first time with the simulation environment »SpaceView & Analyst«.
Efficient sensor and processing networks
The key task of space observation is the measurement of the orbits of objects that circle the Earth to ensure that the position of each individual object can be reasonably estimated at any time. Depending on the traffic area that is to be observed, radar systems or optical sensors can be used to analyze the current situation, represented by a database with the orbital data of the object. Due to a wide range of influences, one data set can only provide high-quality forecasts pertaining to the position of an object for a limited period of time. Hence, the orbits have to be re-measured on a regular basis. Fraunhofer FHR is currently focusing on the development of suitable sensor and processing networks together with support systems for space observation.
Although space observation was partially improved in the past through the utilization of high-performance sensors, the possibilities for optimization and efficiency enhancement at a global level remain widely unexplored. A holistic approach to the process of »space observation« is still lacking. In this respect, the modeling and simulation of the complex interaction between all system components is absolutely essential.
Modeling of complex systems
The visualization and simulation environment »SpaceView & Analyst« was developed at Fraunhofer FHR to facilitate the systematic investigation of space observation systems. The core element is a high-performance module for the representation of the traffic area that is to be analyzed, the objects within this area and the sensors that were employed in terms of place and time. This approach facilitates the modeling of practically all sensor system types, from ground to space-based. The static or dynamic alignment of each sensor can be individually defined. Processing networks and database systems can be integrated in an analogous manner. In this way, strategies, e.g. for the sampling of the observation volume in the respective line of vision of a sensor, or other observation modes, such as beampark or tracking, can be investigated and evaluated.
The software mapping of all components of an observation system that is the subject of investigation takes place on the basis of parallel processes and interprocess communication methods. The simulation capacity is scaled according to the number of processor cores and even facilitates the real-time investigation of complex space observation systems with a large number of sensors and downstream processing units.
Verification using end-to-end simulation
The actual simulation takes place over time: as in reality, the systems involved in the simulation provide measurement data with a temporal resolution that is determined by the respective operating mode and sensor and may also, where appropriate, have to be supplied with data. In the case of a traditional surveillance radar system, the alignment of the antenna is, for example, an unchanging periodic time function. Within a predefined time slot, the sensor captures the objects that are located in the current antenna beam. Once the input data has been filtered at the sensor, the newly acquired observation vectors (position data) are transmitted to a processing system. This system groups the transmitted data with a view to creating groups with data sets which, with a high degree of probability, can be allocated to the same object. A further instance then numerically determines the orbit of the respective object on the basis of this information. The orbital information can then be entered into the orbit database.
Resource-efficient space observation aims to measure the orbits of objects as precisely as possible while at the same time making the most efficient use of all available resources. The quantification of the minimum degree of precision that is required is by no means trivial. Stability criteria, which requires that the number of valid orbital data sets may no longer change after a certain cold-start phase, has to be met. And this of course under the prerequisite that there is a constant number of observed objects, the orbits of which are not actively changed by a maneuver or through a collision with another object. Proof that the stability criteria for a predefined system configuration has been met can only be obtained by carrying out a corresponding end-to-end simulation.