This paper focuses on the development activities of Sodern and Polatis on their Space Optical Switch development both for GEO 1 terabit/s high reliability and for low cost LEO constellations. An overview of the technology evaluations and design studies as well as of breadboard performance and environmental testing will be presented. Finally Sodern will explain the product development roadmap including the upcoming EQM. |
1.INTRODUCTIONThe next generation of telecom satellites is expected to provide high data throughput, while making communication links reconfigurable during all mission phases, with high performance and decreased mass, power and cost per bit. For more than a decade, the space industry has put efforts into demonstrating the advantages of photonics technology to obtain such innovative payloads, with a photonic repeater at the core of the payload, instead of an expensive and complex full RF architecture [5] [6] [7]. The combination of conventional Radio Frequency with the emerging photonic solutions, will thus support flexible routing of any RF signal from any input channel, to any output channel at any frequency. At the core of the repeater, the optical switching is the key enabling photonic technology to bring high flexibility in communication satellite payloads but also to allow redundancy functionalities and optical interconnecting [8] [9]. The optical switching and routing equipment will be responsible for directing incoming optical signals containing Radio Frequency data to suitable output ports. While many of the satellite payload photonic subsystems can be designed based on already space qualified devices, no technology for optical switching has been brought to a sufficient level of maturity to reach the flight proven level within a few years. The terrestrial technology, named DirectLight® [1] [2] [3] based on 1.55 μm fiber matrices, oriented thanks to piezo actuation, turns out to be the best candidate for a space optical switch of large port number. This paper thus describes the space Optical Switch Matrix development, based on the DirectLight®, orientable fiber matrix technology. We first give an overview of the future payload architectures, requiring the optical switch function. We then describe the indisputable advantages of the DirectLight® technology for answering space optical switch requirements of reduced optical loss, mass, power and reliability compromise for a large number of ports. We will then focus on the late advances obtained with the first optical switch breadboard for Geostationary telecom satellite that has been successfully submitted to environmental tests, and finally we will explain the perspectives of the optical switch development either for GEO large satellites or LEO satellites constellation. 2.PHOTONICS PAYLOAD FOR TELECOM SATELLITE2.1Added value and maturity of photonics for communication satellitesThe recent need of GEO telecom satellite capacity and flexibility has drastically increased. The tables below illustrate the expanding demand for geostationary traditional telecom. This implies critical improvement of satellite payloads with regard to equipment numbers, mass, consumption and heat dissipation that the traditional RF technology based telecom payloads is expected to not be able to achieve any more. RF technology has indeed reached its limit of feasibility for price/size/mass launch constraints. Table 1.Examples of Ka band geostationary data services satellites.
Therefore, the main challenge of the next Terabit/s generation of HTS is to provide a ten-fold-increased capacity with enhanced flexibility, while maintaining the overall satellite within a “launchable” volume and mass envelope. In parallel, the emergence of MEO/LEO satellites constellation using laser links enhances the need for full optical signal based telecom satellite with drastically reduced price/size/mass constraints to answer mass production needs. The next table shows the increasing demand for both hybrid or full photonic payloads. Table 2.Examples of MEO/LEO communication satellites constellations.
To that extent, the photonic technology offers an adequate answer to the operator’s interest for higher capacity and payload flexibility but also for full optics communication payload. The next figure illustrates, photonic payloads architecture types with RF data over fiber thanks to optical conversion, or full photonic equipment addressing laser links/and feeder links technologies 2.2Advantages using optical switch for photonics payloadsOne of the main enabling photonic fibered technologies is the optical switching. For GEO satellites, in addition to flexible allocation of gateways, it provides redundancy functionalities and large scale optical interconnecting which is not possible with pure RF technology. To sum up, the switch brings to the payload its flexibility, its versatility and its reconfiguration ability in case of local failure. For LEO constellations, the switch will bring the software defined flexibility that is necessary for routing optical signals at the satellite network scale. The optical switch is one of the cornerstones of the future architecture for photonic payload. This equipment is in charge of distributing and routing of optical signal carried by optical fiber. Optical Switch functions and advantages are:
Sodern, a recognized optronics space equipment provider and HUBER+SUHNER Polatis, the world leader for terrestrial optical switches in many market segments including telecom and data centers, are therefore developing a new space product to answer this need of an optical switch function in future telecom satellites, whatever the kind of photonics based payload (with RF or digital technologies) is required, for GEO or LEO telecommunications. The first efforts of this development have been focused to answer GEO stationary needs, while development for LEO constellations will start before the end of 2018. 3.DIRECTLIGHT® ADAPTATION FOR SPACE OPTICAL SWITCH3.1Directlight® technology presentationNo solution has been developed to high level of TRL for now in Europe for the switch function of the Space Telecom photonic payload. A previous analysis had demonstrated the adequacy of the HUBER+SUHNER Polatis technology with the space optical switch for telecom satellite photonic payload. Polatis terrestrial products address many markets including terrestrial telecom and datacenter, with a core technology capable of providing up to 384x384 ports and beyond. This is then the only identified solution for addressing a large number of ports while reaching 1dB of insertion loss for space applications. The innovative principle of the HUBER+SUHNER Polatis technology DirectLight® [1] [2] [3] is illustrated in the next figure:
The core performance of the DirectlLight® product is based on the slice component composed of 12 fibered collimators with corresponding piezo-actuator and proximity EEE components as shown in the figure below. The integration and alignment of several slices enables to build a complete large switch matrix. 3.2Advantages for space applicationsDirectLight® technology is adapted thanks to its main advantages, interesting for space application:
4.GEOSTATIONNARY OPTICAL SWITCH DEVELOPMENT4.1Technical objectives of the GEO space optical switch developmentSodern and Polatis objectives to develop the space optical switch with DirectLight® technology, are to define a scalable architecture that makes the best compromise between SWaP (Size Weight and Power) and reliability, in order to reach optical performance identical to terrestrial optical switch performance and to guarantee functionalities and performance within the GEO or LEO required environment. Therefore, the major requirements underlying the space Optical Switch development are as follows:
4.2OPTIMA Switch breadboard objectivesThe early development has been funded by an European Commission H2020 Program funding. The project named OPTIMA has the objective to demonstrate and validate the photonics payload concept and its benefits in a relevant industrial environment, which includes delivering a demonstrator of an hybrid RF/photonics payload and the roadmap to such future payload. The OPTIMA consortium gathers European industries piloted by ADS Ltd and constituted of ADS Ltd (UK), ADS SAS(Fr), DAS photonics(Sp), CORDON(It), IMEC(Be), HUBER+SUHNER Polatis (UK) and Sodern(Fr). The project started at the end of 2016 and is planned to be finished by mid-2019. Sodern and Polatis provide the Optical Switch equipment included in the payload demonstrator. At switch level the main objectives in the frame of Optima were then :
The OPTIMA Elegant Breadboard is a 24x24 port switch, constituted of the following sub-systems:
In the scope of OPTIMA, the SDE is the Polatis terrestrial control board, while a specific SCU has been designed based on adapted DirectLight® Slices integrated in a specific mechanical structure, to reach compliance to mechanical environment and interfaces for the OPTIMA payload demonstrator. A folding architecture with mirror facing a set of four 12-port slices has been chosen to reduce the overall size. Though the main effort in the OPTIMA project is focused on addressing slice risk of non-robustness to space, Sodern has evaluated the impact of space grade EEE components on the electronics boards to assess the absence of any showstopper. 4.3DirectLight® technology preliminary evaluationAs explained before, the core performance of the DirectLight® product is based on the slice component (illustrated earlier in Figure 3). The first step of the development to be addressed is the slice compatibility to space constraints either with regards to material and process norms or to its ability to sustain any type of LEO/GEO telecom mission’s environment. Sodern has evaluated the compliance of Polatis Slice with respect to European Space Standards (ECSS) and SODERN Standards applied for the Space Telecom market. A thorough analysis of about 15 Materials, 5 Mechanical Parts and 20 Processes, of the slice manufacturing line (machines, tools, test benches, monitoring too), Electronic Assemblies and soldering processes, Gluing processes, Welding processes has been successfully performed to conclude the compatibility of slice manufacturing with space standards. Sodern has then performed FEM modeling of the slices and an extensive vibration tests campaign on different specimens to identify the slice mechanical weak point, its origin and possible modifications. The main conclusions of slice models and tests correlation is that the terrestrial slice needs modifications to sustain high mechanical environment up to about Random Vibration=26 gRMS at slice interface. A trade-off has shown that the minimum modification concerns its baseplate stiffening and change of its material. However, this sole modification is not sufficient to drastically reduce vibration levels at collimator and piezo level, which will necessitate finding some compensation with the mechanical structure. 4.4OPTIMA Switch Breadboard design approachThe objectives of the Switch Core Unit design, is thus to minimize vibration modes without degrading functionality of the slices and maintain them in an accurate and stable position. The figure below shows the mechanical structure maintaining the set of 4 slices: This design has been optimized via mechanical analyses using a FEM model, to check that there is no issue regarding acceleration on the slices, stresses and screwed loads on the overall Switch Elegant Breadboard, submitted to different loading cases. The main criterion defined during the mechanical tests is the acceleration criteria on the slice limited to 26gRMS random vibrations. High frequency dampers, which are located underneath the SCU as shown in the above figure, enable to limit slice vibration to 26 gRMS with entrance level at switch interface at 16 gRMS. 4.5OPTIMA Switch Breadboard tests resultsAll slices have been produced by Polatis, on its production line, taking into account required modification. Other sub-assemblies – mirror, mechanical structure, dampers – have been procured by Sodern to Polatis for integration of the OPTIMA Switch Elegant Breadboard. Ahead of breadboard integration, some Lot Acceptance Tests on slices have been performed in vibration and TVC, and have successfully passed all functional tests after each environment tests. The integration and calibration of collimator positioning has been performed at Polatis premises in addition to characterization of initial electrical and optical performance (using standard Polatis ambient test facilities), while environmental tests in non-operational mode are performed at Sodern. During this campaign of environmental tests, functional go/no-go tests to check closed-loop collimator positioning performance and insertion loss stability (with a measurement accuracy of ±0.5dB) have been successfully performed between each environmental test. Full performance tests on all port links are only performed at the end (again using Polatis test facilities). Underneath table synthesis main tests results on initial insertion loss on all ports links, and IL stability in between environments on a reduced number of ports links: The graph below focuses on insertion loss performances measured for now: As a preliminary conclusion, integration, calibration and environmental tests have been passed successfully, at slice level and Switch Elegant Breadboard level. At the time of writing, functional tests after TVC are being performed at Sodern, and final performance measurements are planned afterwards back in Polatis with standard Switch bench. At this stage of the breadboard tests, the main conclusions of the first step of space switch development are:
4.6GEO Switch ArchitectureIn parallel with OPTIMA switch bricks validation towards space constraints, Sodern has started defining the Space Optical switch new architecture, both regarding electronics boards and in terms of SWaP optimization. In cooperation with Polatis, the electronics architecture has been analyzed to identify solutions for replacing EEE proximity components of slice or controlling board while maintaining identical functionality and performance. For now, a smart preliminary architecture has been retained without redundancy in order to optimize the overall size and mass compromise. In the figure below, the implementation concept is foreseen with an electronics architecture showing the SDE boards placed around the SCU that enables to have a compact monobloc equipment. The damping function defined in OPTIMA is maintained, while evolution of SCU relies on slice orientation perpendicular to the interface plane. The choice of electronic architecture and overall switch implementation enables Sodern and Polatis to propose a modular concept of switch with a large number of ports and reduced size and mass summarized underneath: Table 3.Foreseen GEO Optical Switch Mechanical interfaces.
Rough orders of magnitude of maximum power consumption have been assessed leading to ranges as follows: 8-24 W for 24x24ports, 17-49 W for 48x48 ports and 33-97 W for 96x96 ports. Preliminary reliability evaluations are also being processed. Though no redundancy has been designed at this stage of preliminary evaluation, the modularity of the foreseen concept will enable to propose necessary redundancy, with the best reliability SWaP compromise proposed above. 5.ROADMAP TO OPTICAL SWITCH FLIGHT MODEL5.1Roadmap to GEO FM optical switchWhile the space telecom industry has been working for years now on introducing photonics in telecom payloads, lately a rapid increase in demand has spurred satellite manufacturers onto working on IOD or ground demonstrators to highlight the assets of photonics. Sodern is now working in cooperation with primes to prepare the switch product for PFM/IOD within a few years. The figure below shows Sodern’s development plan to an Optical Switch flight model. OPTIMA breadboard has confirmed that DirectLight® technology is the most appropriate technology for reaching a high level of reliability and optical performance for large switches. The next step is to produce an Engineering Qualification model, with representative slice electronics and that will be submitted to qualification tests either at component, sub-assembly or EQM levels. The step to an IOD or PFM will depend on satellite primes’ requirements. This roadmap will enable to answer large telecom satellite market for small series of 1 to 5 switches a year. 5.2Opportunities for LEO constellation optical switchSodern and Polatis have proposed to take part to a European Commission H2020 consortium including MDA UK, and DAS photonics. The proposal, named SODaH, plans to consolidate concepts of LEO broadband constellations using OISL and RF Up/Down links enabling an optical space data highway and demonstrate the feasibility of a “Software Defined”, “Hybrid Photonic/Digital” payload. Activities could start before the end of 2018. For the switch function, Sodern and Polatis proposes the same mechanisms as those of the GEO switch: the core technology will be based on the same DirectLight® slice. The proposed use of Screened COTS EEE components and additive layer manufacturing will enable to reach a new range of requirement: an optimized SWaP (2 dm3, 2 kg, 5 W), a low-cost switch for a high volume manufacturing (100s/year) answering future constellation needs. 6.CONCLUSION AND PERSPECTIVESSodern has demonstrated that the HUBER+SUHNER Polatis terrestrial 1.55 μm optical switch, with its core DirectLight® technology can be ruggedized to sustain space environment while maintaining a very low level of insertion loss. First in the scope of the European Commission Horizon 2020 OPTIMA project, Sodern and Polatis have designed, manufactured and tested a 24x24 port Switch Core Unit Breadboard, based on ruggedized slices and an optimized mechanical structure. A test campaign was run both at switch component (slice) level, and at Breadboard level to characterize, adjust and demonstrate robustness of the space design to vibrations up to 16g RMS, shocks up to 1300g SRS and Thermal vacuum cycles up to [-40 °C;+60 °C]. Throughout the environmental tests sequence, the main optical performance, insertion loss, has proved to be stable within ±0.5 dB, with average insertion loss of 1 dB for all 48 ports, which is a very low contributor to the whole future photonic sub-systems for telecom satellites. Those results have brought DirectLight® to a high level of maturity for space application. In parallel, Sodern has also evaluated the DirectLight® controller board adaptation with space norms and space qualified components, to conclude that its space equivalent Switch Driving Electronics will enable Sodern and Polatis to propose a mono-bloc 1.55 μm Optical Switch concept, with a good level of compromise between size, mass and power ranging from 24x24 ports to 96x96 ports for Geostationary telecom satellites. The next step of development for the GEO optical switch is to design and manufacture the 24x24 ports Engineering Qualification Model including space designed Driving Electronics and ruggedized DirectLight® slices in a Switch Core unit benefitting from the OPTIMA Breadboard conclusions. With this new step and further adaptation to future customers’ needs, Sodern and Polatis will be in a position to propose an In Orbit Demonstration of a 24x24 Optical Switch Matrix for GEO satellite within less than 5 years, and get the “flight proven” level paving the road for Sodern and Polatis to develop and produce Optical Switch up to 96x96 ports. Sodern and Polatis are also working on a demonstrator of a low cost DirectLight® based optical switch matrix for new space constellations. It will present significant added value for concepts for LEO broadband constellations using OISL and RF Up/Down links via an Hybrid Photonic/Digital, payload. 7.ACKNOWLEDGMENTSThis work has been supported by the European Commission in the frame of Horizon 2020 (H2020-COMPET-2016 / GA N°730149) The authors would like to express special thanks to the Space Agencies CNES and ESA for support, and their Technical Officers for their advice and encouragement, in particular Laurent Gatet (CNES), Frederic Lacoste (CNES), Laurent Cadiergues (CNES), Frans Ijpelaan(CNES) and Nikos Karafolas (ESA). The authors also thank their partners in the OPTIMA project for fruitful cooperation, especially, with Johan Bauwelinck and Michael Vanhoecke (IMEC), Joyce Wong and Javad Anzalchi (Airbus Defense and Space Ltd UK), Thibault Verges (Airbus Defense and Space SAS Fr), Valentin Polo and Olga Navasquillo (DAS Photonics), and Marcello Tienforti and Antonello Vanucci (Cordon Electronics SRL). REFERENCESN. Parsons, Polatis,
“Dynamic optical circuit switching for efficient and scalable software-defined datacentre networks,”
in Symposium on “ Optical Communications and Networks for Datacenters,
(2015). Google Scholar
N. Parsons, A. Hughes, R. Jensen,
“High Radix All-Optical Switches for Software-Defined Datacentre Networks,”
ECOC, Dusseldorf
(2016). Google Scholar
H. C. Hansen Mulvad, A. Parker, B. King, D. Smith, M. Kovacs, S. Jain, J. Hayes, M. Petrovich, D. J. Richardson, N. Parsons,
“Beam-Steering All-Optical Switch for Multi-Core Fibers,”
OFC, Los Angeles, USA
(2017). Google Scholar
T.Verges,
“Airbus photonics payloads for telecom satellites,”
ARTES ScyLight, ScyLight Industry Day, 2017 ESA ESTEC, Noordwijk, Netherlands Google Scholar
J.Anzalchi1, P. Inigo2, B. Roy2,
“Application of photonics in next generation telecommunication satellites payloads,”
ICOS, Tenerife, Canaries Islands, Spain
(2014). Google Scholar
N.Karafolas,
“Photonics in Satellite Telecommunication Payloads,”
International School on Space Optics, ESA/ESTEC, Netherlands
(2017). Google Scholar
G. Pastore S. Vono, M. Aveline, M. Sotom, A. Pisano, M. Piccinni,
“Roadmap for the introduction of photonic technologies based payloads,”
ARTESDay, January2017). Google Scholar
M. Sotom, M. Aveline, R. Barbaste, B. Benazet, A. Le Kernec, J. Magnaval, M. Picq,
“Flexible Photonic Payload for Broadband Telecom Satellites: from concepts to system demonstrators,”
ICSO, Biarritz, France
(2016). Google Scholar
|