2.1

Digital downlink

A digital downlink is in charge of downloading great volume of space optical telemetry to avoid the RF telemetry bottle-neck. A pseudo random binary sequence (PRBS) mode is also implemented to derive bit error rate (BER) performance (before and after on ground decoder). The downlink optical modulator is implemented in the Terminal Control Electronic (TCE). The generated 10 Gbauds baseband signal is provide at the transmitting Laser Communication Electronic (LCE-TX) input presented in [15] by iXblue. The LCE generates low power non-return to zero differential phase shift keying (NRZ-DPSK) signal around 1560 nm. This low power signal is amplified by the booster amplifier up to 5 W continuous power, described in [16] by CILAS. The On-Board Terminal sends the signal down to the On Ground Terminal. The On Ground Terminal collects the signal and injects it into a single mode fiber interfacing with the receiving Laser Communication Electronic (LCE-RX). The LCE-RX detects the optical signal and provide it to the Optical Modem RX.

Figure 1.

Digital downlink architecture

2.2

Digital uplink

A digital uplink allows deriving BER performance and might allow file uploading in the future. The uplink optical modulator is implemented in the Optical Modem TX made by Airbus Defence and Space. The generated 10 Gbauds baseband signal is provide at the on-ground LCE-TX input designed by iXblue. The LCE-TX generates low power non-return to zero on-off keying (NRZ-OOK) signal around 1550 nm. This low power signal is amplified by a ground booster amplifier up to 10 W and maybe 30 W continuous power. The On Ground Terminal sends the signal up to the On Board Terminal. The On Board Terminal collects the signal and injects it into a single mode fiber interfacing with the LCE-RX presented in [15] by iXblue. The LCE-RX detects the optical signal and provides it to the uplink demodulator designed and implemented by Airbus Defence and Space in the TCE.

Figure 2.

Digital uplink architecture

2.3

Analog uplink

An analog uplink is also implemented for a transparent optical feeder link proof of concept for broadcast system and to allow uplink carrier to noise ratio estimation on-board the satellite at the LCE-RX output with around 0.5 dB accuracy. The Software Defined Radio (SDR) TX modem generates a 40 MHz large DVB-S2 signal in L-band which is then transposed around 1550 nm by amplitude modulation (reduced carrier single side band or dual side band) thanks to the Analog Modbox TX. The resulting signal goes through the booster, the on-ground terminal, the on-board terminal and the LCE-RX like the digital uplink signal. At the terminal control electronic, the 40 MHz DVB-S2 signal is digitalized, encapsulated and sent back to the on-ground SDR DVB-S2 modem RX. The overall link from/to SDR DVB-S2 MODEM is protected by an upper layer forward error code (FEC) designed by Airbus Defence and Space. The concept of upper layer FEC to protect video transmission has already been studied [16]. Nevertheless, it would be the first end to end demonstration up to our knowledge.

Figure 3.

Analog uplink architecture

3.1

Global description

The optical on-board terminal (OBT) is composed of several parts:

3.2

Uplink received irradiance

The uplink irradiance power can be monitored by the TOP-M acquisition and tracking sensor at a frequency superior to 1 kHz. The accuracy is expected to be better than 1 dB over a range greater than 20 dB.

3.3

Uplink injected power monitoring

The uplink injected power can be monitored by the on-board LCE-RX at a frequency of 20 kHz. Such feature has been tested under relevant power time series injected into the LCE-RX fiber for a GEO feeder link pre-compensated uplink. Figure 4 depicts the power error distribution of the estimator. It shows that more than 93% of the time series can be retrieved with error less than 0.5 dB.

Figure 4.

Absolute error on power injected into the LCE-RX for a worst case uplink power fluctuation

4.1

French OGS (FrOGS)

In the frame of DYSCO project, the FrOGS consortium (CNES, and the companies, Safran Data System, OGS Technologies, ALPAO, and Airbus Defence and Space) is currently developing an optical ground station (OGS) for demonstrating optical communications in free space and, more particularly adapted to the cases of GEO feeder links and TMI-LEO downlinks. This development is supported by CNES and co-financed by the aforementioned manufacturers. FrOGS will be located at Observatoire de la Côte d’Azur, Calerne.

The OGS is based on a bi-static architecture with an Alt-Az mount supporting both the TX laser emission optics and the RX satellite stream reception optics (Figure 5 and Figure 6). The transmitting and receiving optics are compact 500 mm diameter telescopes designed with a rigid, light and stable mechanical structure.

Figure 5:

Schematic architecture of the FrOGS station, developed by the French consortium (CNES, ADS, ALPAO, OGS Technologieset SDS)

Figure 6.

CAO view of the instrument optics on the Alt-Az mount

The focal instrumentation of each optical tube consists of a Pointing-Acquisition-Tracking (PAT) system, a fine-pointing device (tip-tilt mirror) and high-resolution adaptive optics (AO) system. The AO system is based on a deformable mirror and a Shack-Hartmann wave front analyzer. Each adaptive optics is independently slaved to the RX optical stream received by the two telescopes. The TX optical setting also includes a forward pointing system based on a second tip-tilt mirror allowing for the satellite velocity aberration compensation.

The RX focal instrumentation is directly coupled at the RX telescope Cassegrain focus and embarked behind the primary mirror. The opto-mechanical interface at the focal RX instrumentation output is designed to allow coupling with a single-mode optical amplifier for GEO feeder links or with a detector in free space for the TMI-LEO links.

The TX instrumentation is located on the lower floor in an environmentally-controlled focal laboratory. It connects the TX upstream generated by a 43 dBm power fibered amplifier (booster) with the TX telescope through a motorized coudé optical train.

In parallel to the TX and RX optics, the OGT incorporates two 100 mm bezels, respectively for beacon functions (continuous laser diode amplified by a 40 dBm fiber amplifier) and wide-field sensor (visible and infrared imaging sensor).

The telescope mount is driven by DirectDrive motors combining speed for LEO tracking and pointing stability at the μrad level. The entire OGS is protected by a slit dome on the roof of a 4 x 4 m size building. The station can be operated locally or remotely through a secure link.

The baseline architecture of the demonstration carried out in the frame of the DYSCO project relies on this French OGS in order to assess its ability to efficiently mitigate atmospheric effects on a long-term basis. This demonstration prototype should be a precursor of further operational OGS.

4.2

Airbus OGS (A-OGS)

Airbus Netherlands is providing an OGS solution for GEO feeder link demonstration via the ScyLight activity CREOLA. This solution is split into three phases, in a step-by-step approach. In Phase 1, Airbus will operate the already validated ESA OGS with support of its partner GA-Synopta. In Phase 2, Airbus will add its own adaptive optics corrected TX-OGS in a transportable configuration, together with partners ASA and TNO. Combined with ESA OGS, a bi-directional feeder link is then operated. In phase 3, Airbus will demonstrate a bi-directional feeder link with a monostatic OGS, based on an RX/TX AO system from partner GA-Synopta.

4.2.1

Short term bi-static approach

The chosen bi-static architecture allows for a gradual growth in link complexity. OGS key technologies will be demonstrated in genuine space-ground link conditions, providing valuable feedback for OGS product designs.

An important objective is to validate the predicted efficiency of full-aperture RX-AO and pre-compensated TX-AO on a 20 cm uplink communications beam in the 1550 nm band. Over a genuine GEO satellite atmospheric channel, tests in different atmospheric conditions will lead to a comprehensive optical channel analysis.

Local atmospheric turbulence at the telescope level is considered as a major cause that disturbs an AO system performance. Those turbulence are induced by surface layer’s flow distortions and local thermal turbulence from nearby infrastructure elements. Predictions on associated AO performance degradation for a transportable OGS, with a large aperture telescope in a tight dome enclosure, do not exist yet. Phase 2 collects measurements to characterize local turbulence contributors and their related AO impact, to validate key parameters of future OGS products.

Figure 7.

Bi-static OGS approach for Phases 1 & 2

4.2.2

Monostatic OGS prototype

The monostatic OGS (M-OGS), carried out in the third phase, constitutes a platform that provides all functions for bi-directional GEO optical feeder link testing.

To minimize transport overhead, the M-OGS will be tested in a stationary OGS environment at Zimmerwald site location in Switzerland. At a lower site altitude, but at almost identical zenith angle to TELEO, this allows for a comparison of fundamental findings from Phases 1 and -2, collected at Izana site on Tenerife.

The M-OGS provides full aperture RX-AO over 80 cm, combined with AO/based TX-pre-compensation for two TX beams of 20 cm diameter. The AO module form factor allows a later integration in a transportable container, like demonstrated in Phase 2.

The M-OGS will re-use the external beacon module from Airbus NL, and it will implement parts of the OGS Control and Safety System, already tested during in-orbit demonstration phase 1.

Figure 8.

Monostatic OGS for Phase 3

5.1

Space segment

5.1.2

Laser Communication Electronic

Figure 9 presents a picture of the LCE demonstration model.

Figure 9.

Picture of space grade LCE

The LCE allows the conversion of the electrical data into optical data and vice-versa. It interfaces electrically with the DPU part of the TCE and optically with the focal plane assembly and the booster. It is managed directly by the TCE.

The LCE is composed of three electronic boards:

• ● DC/DC converter board. It aims at converting the primary unregulated 100 V bus line distributed by the GEO satellite into specific secondary voltages required by the transceiver. It is composed of a front-end generic converter and a back-end specific converter for modularity.

• ● TX board. Its role is to convert 10 Gb/s NRZ digital electrical data into 10 Gb/s NRZ-DPSK optical signal. The board embarks two optical channel emitters (OCE), one nominal and one in cold redundancy. The OCE is a fully packaged sub-assembly composed of:

  • ○ A continuous wave (CW) distributed feedback (DFB) laser, thermally regulated by a thermo-electrical cooler (TEC) and generating the optical carrier locked at a specific wavelength. The package also includes a monitoring photodiode (PD)

  • ○ A Mach-Zehnder modulator (MZM), modulating the optical carrier

  • ○ A RF driver, composed of three amplification stages for DPSK modulation, able to amplify the signal coming from DPU up to the required voltage for the MZM

    Both OCE optical outputs are multiplexed into a single fiber using an optical coupler. After that, an optical coupler and a shared monitoring photodiode (over the two channels) allows picking up small amount of optical power to perform the modulator bias loop control.

• ● RX board. It amplifies the weak optical signal captured by the telescope and injected into the receiver fiber and perform the opto to electrical conversion. It is composed of a low noise optical amplifier (LNOA), an all-fibered demultiplexer and three optical channel receivers (OCR). OCR are photoreceivers able to demodulated both OOK digital optical signals and intensity modulated analog optical signals depending on their operating gain parameter. One OCR is dedicated to demodulate 10 Gb/s NRZ-OOK optical signal and the two other OCR, one nominal and one in cold redundancy, are devoted to demodulate analog optical signals with a bandwidth from few hundreds of kHz up to 23 GHz.

  LNOA includes also an input monitoring photodiode able to estimate the optical power injected from the focal plane assembly into the optical communication chain receiving fiber with a 10 kHz bandwidth and a dynamic range around 25 dB. This will allow estimating the uplink turbulence effect on the optical power injected into the optical communication chain.

Basically the specification of the LCE unit are presented in Table 1.

Table 1.

On-board LCE main features

Feature	Specification
Power bus input	● 100 V bus
Controllability	● CAN, analog telemetries
RF input	● 10 Gb/s NRZ electrical signal Single ended 150 mVpp min eye amplitude
TX optical output	● Single mode polarization maintaining fiber
● 10 Gb/s NRZ-DPSK optical signal
RX optical input	● Single mode fiber
● 10 Gb/s NRZ-OOK optical signal
● 500 kHz to 23 GHz analog bandwidth
RF output	● 10 Gb/s NRZ electrical signal Differential 1000 mVpp max eye amplitude
● Analog electrical signal Single ended
Wavelength plan	● OCE1: 1562.64 nm nominal channel
● OCE2: 1561.01 nm backup channel
● OCR1: 1550.12 nm backup analog channel
● OCR2: 1548.51 nm nominal analog channel
● OCR3: 1546.92 nm nominal digital channel
RX input optical power monitoring	● Range: 25 dB
● Bandwidth: 10 kHz
Electrical power consumption	● For all OCE and all OCR: 27.9 W
● For 1x OCE and 1x OCR: 21.3 W

More information can be found in [15].

5.1.3

Laser Power Electronic

Figure 11 presents a picture of the LPE demonstration model.

Figure 10.

Picture of the space grade LPE

Figure 11:

Picture of the Lasercom modem

The booster is composed of a main electronic board embarking the DC/DC converter and the control processor, an opto-electronic mezzanine and a baseplate onto high dissipative elements are reported.

Optical amplification up to 5 W continuous power with wall-plug efficiency of around 11% (beginning of life) is based on two amplification stages:

• ● Stage 1: pre-amplifier. Low power single mode pump diodes are used for this stage. This amplification stage aims at shaping the optical spectrum for the second amplification stage and increasing the signal power for power-amplification.

• ● Stage 2: power-amplifier. High power multi-mode pump diodes are used for this stage. This amplification stage aims at increasing the signal power up to the required power with required noise factor and providing good spectrum flatness.

Three monitoring photodiodes (PD) are used to monitor the optical signal and to secure the booster operation (input power, inter-stage power and output power). Thermistors are also used to monitor temperature on different components in order to protect them against overheating.

The booster communicates directly with the TCE through CAN link.

Table 2 presents the main specifications of the on-board booster.

Table 2.

On-board LPE main features

Feature	Specification
Power bus input	● 100 V bus
Controllability	● CAN, analog telemetries
Optical input	● Min: 6 dBm
Operation wavelength	● From 1555.7 nm to 1565.5 nm
Output power	● From 0 up to 5 W continuous power
● Stability <±0.3%
Wall plug efficiency	● 11.2% at 5 W BOL

More information can be found in [16].

5.2

Ground segment

5.2.2

Optical Modem RX

Figure 12 presents a picture of the Lasercom digital processing unit. The lasercom Rx multi-mission modem performs digital processing from Rx photodetector output up to payload bits delivery on TCP/IP. In the framework of this paper the lasercom modem processes the RF output of the LCE-Rx and delivers payload bits on 10Gbps Ethernet and on 10TB HDD.

Figure 12:

Test bench for digital downlink

Figure 13 :

Cortex Lasercom overview

The lasercom is composed of 4 main parts (see figure 13):

• 1. Cortex HDR server. This part is composed of a 4U chassis equipped with a server board driving DSP boards and providing legacy HDR FTP and TCP/IP interfaces.

• 2. Acquisition board. The role of this board is to digitized the 10Gbps DPSK RF signal converted by the photodetector (LCE-Rx).

• 3. Signal Processing board. This board performs the following functions:

  • a. Resampling and matched filtering

  • b. Clock recovery

  • c. Despreading

  • d. LLR computation fitting any OFE (EDFA+PIN, APD, …)

  • e. Deinterleaving up to 8GB

  • f. Provide channel metrics synchronized with demodulated data: ROP, SNRe, electrical level.

• 4. FEC board. This board is dedicated to channel decoding and data integrity. In the framework of this paper, it performs QC-LDPC decoding, BCH and CRC checking.

6.1

Digital downlink

Figure 12 presents the test bench for the digital downlink scenario. The LCE-TX and booster are electrically powered by 100 V lab source. They are controlled through graphical user interfaces on a computer with CAN and UART probes. TCE-EM realizes the FEC encoding, interleaving and framing. At the receiver side, the LCE-RX is used to demodulate the optical signal which is then transmitted to the optical modem RX. The Optical Modem RX-EM performs synchronization, deframing, desinterleaving, and decoding in order to retrieve the initial useful data.

6.2

Digital uplink

Figure 13 presents the test bench for the digital uplink scenario. On the on-ground side, the DPU-TX generates either PRBS 2⁶³-1 or encoded and interleaved data to be transmitted. The 10 Gb/s raw NRZ signal modulates the on-ground optical transmitter breadboard. The LCE-TX output is a 10 Gbaud NRZ-OOK signal at 1546.92 nm. A commercially of the shelf optical booster is used to amplify this signal up to 10 W continuous power. The received signal is demodulated by OCR3 of LCE and converted back into electrical signal processed by the TCE DPU board.

Figure 13:

Test bench for digital uplink

7.1

Bit error rate before decoding

Uncoded downlink bit error rate (BER) versus received optical power (ROP) are presented on Figure 14. Uncoded data is based on pseudo-random binary sequence (PRBS) of length 2⁶³-1.

Figure 14:

Uncoded downlink BER for 5 W booster output power, 10Gbauds, NRZ-DPSK

Both downlink channels have the same performance and the overall communication chain presents a state-of-the-art sensitivity of -45 dBm for a BER of 10^-3.

Uplink uncoded BER versus ROP for 10 W booster output is shown on Figure 15. An overall communication chain sensitivity of -41 dBm is reached for BER of 10^-3.

Figure 15:

Uplink uncoded BER for 10 W booster output, 10Gbauds, NRZ-OOK

7.2

Frame error rate after decoding

7.2.1

Digital downlink

Figure 16 depicts the FER after decoding for the new LDPC codes presented in [19]. It shows that simulator in floating point and measurement on test bed with the hardware presented in the previous sections § 5 are aligned with 0.3 dB except for the MODCOD with a FEC ratio of 3/10 and a spread factor (SF), also called repeat factor, of 4.

Figure 16.

Downlink FER for 5 W booster output power

7.2.2

Digital Uplink

Figure 15 depicts the FER after decoding for the QC-LDPC codes presented in [20]. The useful data rate is limited on the uplink with a repeat factor of 8 due to TCE resources limitations. A quasi-error free FER (FER < 10^-6) is obtained for ROP of -50 dBm with QC-LDPC codes R = 1/2 and SF = 8.

Figure 17.

Uplink FER for 10 W booster output power