The development of commercial operators led to new approaches to improve the productivity of infrastructures in space and on the ground to increase profitability. The introduction of software-defined technologies is considered a major technological trend to reduce costs and improve the efficiency of operations. On one hand, the shift from hardware functions to software enables mass and cost reduction. On another hand, software solutions support the automatization of operations and make systems more flexible and scalable as configurable with a simple file upload. This article will review the potential of software-defined technologies to leverage new opportunities for Telecom, Earth Observation, and Defense applications but also challenge the current value chain.
Telecom: Configurable payload and integration with terrestrial network.
Software-defined satellites are raising interest from Satcom operators who are interested in more flexible communication to better address demand for connectivity. Traditionally, a GEO satellite was launched for a specific mission with an essentially fixed design during its whole life (i.e. 10-20 years). Once in orbit, it could not repurpose its mission even if the demand changed. A flexible satellite uses a Software-defined(SD) payload to reconfigure the antenna beam on-demand that is programmable by sending a new program in uplink communication. The term of “Software-defined” itself is not perfectly defined yet because it can cover a set of different design making the term? not perfectly defined yet. We refer to “Software-designed” as any traditionally implemented hardware components replaced by software to configure a function dynamically and programmatically. Essentially a software-defined satellite should offer the ability to dynamically modify the coverage beams and the capacity and power distribution.
This means the mission of satellite can vary during its lifetime depending on demand dynamics. It is particularly suited for the mobility market as the beam can provide coverage to moving targets such as aircraft or vessels or to cover short temporary events (e.g., natural disasters, remote operations, exceptional high demand for communication, etc.). SDR solutions can potentially increase communication reliability by adjusting the frequency in jamming areas or adapt satellite solutions to highly dynamic terrestrial competition or new frequency regulations in a country. On the manufacturing side, this new type of satellite is anticipated to create new demand for communication systems while optimizing the GEO orbit with the replacement of several satellites by one multi-mission satellite.
So far, only the recently launched Eutelsat Quantum hosts a fully reprogrammable payload based on SD-technologies. Two SD-satellites from Inmarsat should follow this year with steerable spot beams and dynamic power allocation. With the exception of some military systems, the development of software-defined satellite is essentially driven by commercial operators willing to provide flexible communication services. The civil government expects a slower adoption, starting with commercial services and potentially ordering proprietary systems in the second part of the decade.
In order to consider an end-to-end software defined and optimized network, the ground segment also needs to adapt to the new requirement in fleet and capacity management. A software-defined solution on the ground can be seen as a way forward to align gateways with the changing communication requirements of a GEO flexible satellite and the management of the increasing traffic. Indeed, the satcom industry being increasingly data centric with more than 9 Tbps of traffic to manage by the end of the decade with a 5-fold increase compared to 2020.
Within the same NGSO constellation, communication links between satellites may differ due to incremental innovation by batch. SD provides the ability to communicate simultaneously with satellites with different RF links, mitigating costs related to communication improvements. For instance, as Euroconsult anticipates more constellations to shift from Ka-band to Q/V-band in the future, SD solutions would mitigate costs due to transition. SD technologies are also easier and quicker to maintain or upgrade, resulting in shorter times of interruption. This time reduction enhances the Quality of the Service (QoS) due to better service continuity and availability.
Managing the network of a NGSO constellation is a complex task due to the hundreds of ground stations to coordinate. Such a network has to be automatized and centralized for effective operations. A Software-defined Network (SDN) enables a dynamic and programmable network through a software interface. The concept of SDN is already well-established in the world of terrestrial networks. It is often used to make different networks interoperable to create one unique virtual network. The introduction of SDN technologies in the space sector could improve satellite integration with terrestrial networks. This interoperability is considered essential to include satellite communication in the 5G transition and boost the Satcom demand.
Earth Observation: reducing data cost and latency
The EO industry is also operating a major transformation aiming to generalize EO satellite-based services to more end-users and expand existing sectors. Cost of data and temporal resolution are often considered the main inhibitors to increase the adoption of EO solutions. The EO industry is then shifting toward more flexible and productive infrastructure. More operators are then miniaturizing their payload to adopt a constellation approach and offer cost-effective high revisit. One particularity in EO is equipment is not used at full capacity due to the passing nature of EO satellites. More ground networks are then being multi-mission to mutualize antenna capacity and improve ground systems productivity.
However, a multi-mission network requires ground equipment compatible with all the satellites to address, which often results in a multiplication of the number of RF chains. Therefore, these multi-mission ground stations can make the virtual management even more relevant than in Satcom. Ground operators can instead virtualize RF functions and digitize RF signal as close as possible to the antenna to mutualize RF resources through an SDR (software-defined radio). Monitoring a constellation also requires a fleet of ground stations working together in the same network far away from each other. A SDN improves the integration of these ground stations by centralizing network management and enabling a dynamic and programmable network.
Virtualization is particularly relevant for the Ground segment as a Service (GSaaS) providers who need to manage a large network of ground stations and support various types of missions. In addition to a centralized and dynamic network management system, Software-defined technologies provide business scalability to these companies by facilitating the integration of new ground stations when the demand increases and to limit the customer acquisition cost. Virtualization will then support the expansion of GSaaS networks. The number of ground stations owned by GSaaS companies is expected to double by 2030.
For instance, KSAT developed the KSAT Lite with a network centric approach rather than the traditional station centric approach. The network aims to address new constellation issues and can be managed through a common interface. AWS developed virtual functions on its proprietary cloud to operate its network of antennas. Microsoft integrates Kratos’ OpenSpace tools in its Azure Orbital cloud-based GSaaS service. RBC signals and Infostellar offer cloud-based protocols to connect any third-party ground station to their virtual network. Space agencies are also interested in virtualization to improve the integration of commercial satellites with national capability and support academic smallsats. Some commercial satellite operators are developing the “Satellite as a Service” concept where a client can temporarily control a satellite through an SDN-based virtual ground station.
Shifting most functions between the antenna and end-users hands – including data management and analytics tools - certainly improves the processing workflow and reduces service latency. Some companies like Exodus Orbitals want to go further by performing analytics directly onboard the satellite to reduce operations on the ground. Through this new software-defined satellite concept, operators can reconfigure the satellite at any time by uploading a new program from the ground and downlink only the data of interest – while current satellites also transmit failed acquisition due to cloud coverage, for instance.
Defense: Reaching permanent coverage through interoperability.
One of the key challenges for military communications is to keep forces connected anywhere at any time. Therefore, defense organizations need to complement their proprietary military system with commercial systems like Intelsat, Inmarsat, SES or Eutelsat. As of today, most military ground systems use closed and protected hardware components designed to communicate with a single or a limited number of satellites, resulting in a low interoperability between systems.
NGSO constellations and their promise to reach global coverage are raising military interest. However, the current lack of flexibility of systems requires a transformation toward more open and integrated ground architectures to switch between constellations to prevent jamming or unavailability. Therefore, defense organizations are considering IP-based solutions to leverage the potential of SDN through a cloud. For instance, the U.S. Space Force requested a budget of $43 million for its fiscal year 2021 budget to develop and test an integrated architecture for commercial and military satellite systems called the “Fighting SATCOM Enterprise”.
IP-based communication will likely change the way to communicate in the military world but will also come with cybersecurity issues. Indeed, the use of this protocol and associated complex architectures might increase the risk of different attacks by increasing the entry points (e.g., denial of access, masquerade, backdoors, malware, spoofing, eavesdropping). The U.K. Skynet 6 and the French Syracuse 4 satellites will include IP capabilities to enable interoperability between commercial or allies’ systems.
A similar trend might be expected for IMINT (Imagery Intelligence) applications as governments are experimenting with the potential of EO high-revisit systems. In the U.S., the NRO and the USAF – who used to procure exclusively Maxar imagery as commercial source – are looking at ways to integrate data from new U.S. operators through study contracts (e.g. as Planet, Blacksky, HawkEye360, HyspecIQ and Capella Space). Software-defined technologies might be an enabler to reduce the gap for connectivity or surveillance and boost commercial use of Satcom and EO data.
Software to disrupt hardware?
On the ground segment, ground stations for NGSO satellites are expected to be the early adopters of software-defined technologies likely the most addressable software-defined technologies representing a total addressable market of about 5,000 operational stations by 2030, representing only 10% of the overall market in units. A large market should then remain for hardware manufacturers. On the space segment, over the 90 satellites identified in backlog about 30 satellites are considered as flexible (but not necessarily software-defined). With the approximately 260 satellites forecasted by Euroconsult to be launched in the next decade, we can expect more ordered flexible satellites in the next decade.
Digitization of the space and ground segment will be transformative for the ecosystem, from hardware manufacturers to solution providers. We are still in the early days of this transformation. Software solutions are anticipated to gain traction in the future with cloud providers such as AWS and Microsoft taking advantage of the virtualization. However, not all RF components can be virtualized easily. The closer to the antenna the more challenging to digitize.
Read more from Euroconsult.