Skip to main content

Satellite Swarm Sensor NETwork

Periodic Reporting for period 2 - S3NET (Satellite Swarm Sensor NETwork)

Reporting period: 2017-05-01 to 2018-10-31

Regarding EO from space, modern approaches show a trend towards fractionated and distributed sensor missions, possibly carrying different types of sensors acting in a formation or swarm. Such missions promise enhanced imaging and service quality and a decrease of deployment costs for satellites due to smaller size and less complex requirements.
S3NET worked on key enablers for development of satellites swarm through optimised and enhanced use of on-board resources.
Five objectives had been defined:
OB1: Reduction of satellite costs (enhancement of mission planning)
Demonstrate through results of S3NET benchmarking studies on “core service applications” and overall effectiveness study, the advantages satellite swarms can bring to a variety of NRT EO applications based on distributed and fractionated satellite sensors.
OB2: Mission scalability and incremental deployment: Achieve breakthrough progress in flight algorithms by proposing cluster control methods (including inter-satellite communication).
OB 3: Advances in performance of the quality of service: Reduce the time from data acquisition to delivery to the end-user.
OB4: Improvements payload processing performance: Enhance knowledge and demonstrate possibilities of high-performance payload-processing systems through the design and implementing of the S3NET concept compute system.
OB5: Coherency between measurement sources: Develop a communication and networking simulation system to enable interconnection approaches and greater coherency of different measurement sources for fractionated and distributed EO applications. Based also on communication enhancements of satellite on-board communication platforms S3NET aims for an increase in data (payload and synchronization data) transmission.
Towards OB1
A dependency analysis that connects costs with technical aspects has been derived from mission parameters. Scenarios, related to 2 optical and 3 SAR fractionated scientific payloads have been analysed in terms of trajectory design, station keeping and platform design to quantify their feasibility. Two main actions contribute to reduction of satellite costs: The fractionation approach and the mass reduction. Fractionation of a mission assists reducing costs for non-recurring developments compared to monolithic missions. Due to strong correlation of costs to mass, miniaturization of on-board processing can reduce size and power – therefore mass and costs.
Towards OB2
S3NET defined four categories of satellites clusters that differ from each other in the number of required orbital planes, inter-satellite distance and tolerances, in order to match different types of mission use-cases. For each of the 3 cluster types a cluster control algorithm was developed in order to minimize fuel consumption, required computational power and communication bandwidth. The algorithms were designed to work with any number of satellites to limit the effect of incremental deployment on the cluster performance. The research emphasized on usage of natural forces such as drag, solar radiation, and Earth geopotential as a mean to reduce the required ΔV. S3NET results are currently used in a collaboration between Technion and ONRG and will be part of future FF missions.
Towards OB3
A subset of SAR image formation & optical pre-processing capabilities were implemented onto the S3NET Concept Compute System. For the SAR case, additional slight multi-looking is mandatory in order to further reduce the amount of data for downlink compared to BAQ. For the optical case, cloud masking and CCSDS-123 compression have been selected to reduce data downlink. Due to new protocols like CFDP and DDS final data products generated onboard can be downlinked to the final user at high data rates and low latency for larger satellite swarms.
Towards OB4
S3Net has analysed technical requirements for a concept on-board compute system and demonstrated high-speed on-board SpaceFibre communication. Also, several benchmarks were demonstrated using RC64 ITAR free, radiation tolerant processors and COTS. Furthermore, S3Net has delivered high-performance software kernels that allow faster processing than under utilization of even open-source optimized software.
Towards OB5
S3NET enhanced the effective data rate within inter-satellite links using a proper multiple access technique. Considering requirements regarding data latency and flexibility (ad-hoc) of the network user separation by time, frequency or code is beneficial. Especially in applications with fractionated sensors a very small delay or precise synchronisation (time stamp) is necessary.
The S3NET communication middleware DDS supports various network topologies by implementing a data-centric, loosely coupled communication system with global data space, and dynamic discovery and improves data timeliness, availability and resource usage.
Formation Flying (FF)
Although recent FF research did not reach the commercial domain, the complexity of large swarm of autonomous satellites deterrents commercial companies. S3NET results offer new and simple design approaches reducing the complexity of the FF. Utilizing natural forces to reduce fuel consumption can reduce costs, and making FF technology more accessible can reduce mission costs.
Optical
In S3NET, small and inexpensive sensor missions in LEO orbit for EO were studied, in order to select representative mission scenarios to investigate in terms of applications. Specific methods to process satellite data were defined and algorithms were developed to cope with the user needs. Co-registration methods, applied to multi-scale and multi sensor data, on-board data reduction through cloudy pixels identification and data compression techniques were analysed. Enhancing on-board performance of satellites and algorithms were developed.
SAR
In the future, space borne SAR missions will fly in close formation, using companion receive-only satellites for (single-pass) interferometric sensing or splitting/extending their synthetic aperture among different satellites. Onboard image formation was investigated, to create realistic benchmarks and determine feasibility. For bistatic SAR imaging employing companion satellite synchronisation and image formation algorithms were developed and integrated into a bistatic SAR scene simulator, a valuable tool for assessment of future mission concepts intended to enlarge the number of scientific observables of SAR missions.
Computing
S3NET firstly created a scalable software-managed high-performance, low-power hardware system employable in space, comprising of multiple high-performance space grade DSPs and off-the-shelf space-validated processors. Secondly, S3Net created a demonstration of software applications and automatically tuned software kernels, for determination of efficient utilization of hardware. This will cause a shift from processing in ground stations to on-board processing due to major cost savings.
Communication
Due to challenging requirements on flexibility/versatility in swarm structures, regarding e.g. number of satellites, topology with a master satellite, distance between satellites, static/dynamic constellations, in principle all parameters relevant to communications schemes were studied.
Adoption of approved techniques from terrestrial cellular networks also wireless sensor networks offered promising solutions to ensure a high quality of service even for bigger and/or ad-hoc satellite swarms.
Project Logo
Concept compute system