NASA/GSFC has played a leading role in the development of direct detection (DD) optical communications for space applications. Because optical technology promises to support Gbps data rate channels with modest telescope diameters of 20 to 30 cm, it is a prime candidate for satisfying emerging high data rate relay needs of NASA and the commercial communications satellite industry. Relevant NASA missions include the Earth Observation System (Eos) and Space Station Freedom, as well as the longer term "Mission to Planet Earth" and lunar exploration. In the commercial world, optical intersatellite links (ISLs) between geostationary satellites would enable a major expansion and improvement in communication services.

A theoretical investigation on optical transmission schemes for free-space applications is carried out. Receiver sensitivity of direct detection schemes like NRZ and QPPM (quaternary pulse position modulation) are evaluated. For bit rates between 10 and 65 Mbit/s, QPPM outperforms NRZ by up to 2.0 dB.

The pointing, acquisition and tracking (PAT) subsystem is a very substantial part of a free-space optical communication system. The requirements, strategies and performances of the PAT functions are presented, together with their influence on the link quality.

In the optical communications frame, the telescopes (emitter and receiver) are used as optical antennas they must have a high optical quality, low central obscuration, low straylight, low weight and low overall dimensions. We are studying, in the SILEX program, the telescope which is an afocal system composed of a Ritchey-Chretien telescope and a dioptric collimator. The parabolic primary mirror (250 mm diameter) is lightened to reduce weight. The mechanical concept is to join the periphery of the secondary mirror spider to the periphery of the primary mirror by a carbon fibers truss which permits high stability. The whole sets of supports are fixed on a sandwich base plate. Thermal environment and operating conditions lead to insure radiative uncoupling between mirrors, truss and outside of telescope and to use thermostated heaters on structural parts. The straylight, mainly due to the sun, can be limited by baffling systems and appropriate choice of materials (studied by MBB). As the telescope optical antenna gain depends on optical quality, optical transmission, obscuration and off-axis effects all these performances must be preserved during the whole flight. We present the study of this telescope whereas a breadboard manufacture and tests are in progress in BERTIN.

An optical inter-satellite communication link should be capable to operate even when the Sun comes close to the pointing direction of one terminal. Therefore straylight suppression at small Sun aspect angles is one of the topics in the Tx/Rx optics design. Each terminal should also be capable to simultaneously transmit and receive laser signals. Therefore, backreflection of transmitted laser light into the receiving channel also has to be care-fully investigated.

The multi/demultiplexers use filtering, polarizing and phase retarding means to combine and split 4 optical channels, spectrally close to each other (821, 827, 833, 839 nm) with an high isolation ratio. The spectral isolator using similar means shows 3 paths : transmitting, communication and stracking. In nominal mode it transmits the MUX channels, receives the 870 nm communication channel and divides it into a given apportionment ratio to provide the tracking function. In calibration mode it connects the transmit path (MUX) to the tracking path. The performances of these equip ments require to quality optical components and technologies (coatings, polishing, assembling...). Part of these activities have been funded by CNES contract.

To meet specifications that are required for demultiplexing in optical telecommunication, we are interested in complex multilayer systems with narrow pass band and high rejection. For spatial application, these filters must have very good optical and mechanical properties and we are led to use hard materials such as oxides. However, these materials deposited with classical vacuum evaporation-condensation techniques have a refractive index highly sensitive to atmospheric conditions. Optical properties of these systems are then very different in vacuum and in air. We show here that a solution consists in using new deposition techniques such as Ion Assisted Deposition. Some examples of multilayers produced with this deposition technique are given and optical performances that we reach are pratically independant of thermodynamical surrounding conditions.

In the frame of Silex project, a prototype model of the Acquisition Sensor (AS) was realized by Officine Galileo; scope of this sensor is to detect the optical signal transmitted by another terminal (either on-ground or on-satellite), and to measure the two-axes position of the incident spot beam. The optical system for focusing the beam on the detector is not part of the AS itself. The sensor is split into two units: a Detection Unit (DU), containing the CCD detector and the proximity electronic, and the Electronic Unit (EU) containing the processing electronic that drives the DU and evaluates the coordinates of the spot focused on the detector. The first unit was a flight-oriented realization; an accurate positioning procedure for the detector (few micrometers of error) was developed. The capability of the sensor to operate with minimum arrangements in the wide range of conditions imposed by the different links GEO-GEO and LEO-GEO is investigated. The most critical aspect, the choice of the detection threshold at varying of input signal power, background noise and operating frequency, is well explained. Finally, results of preliminary tests are reported.

The THX 31160 is a small size sensor which has been designed for optical intersatellite links* . The performance characteristics of the device (image rate up to 20 kHz and improved response in the 0.8 - 0.9 μm range) have been optimized to achieve e very high pointing accuracy of the tracking system. Its characteristics make the THX 31160 a direct competitor to the classical 4 quadrant detector for pointing systems.

Optical communications between satellites in orbit requires highly accurate tracking of incoming laser beam, to allow the generation of necessary control signals for the pointing mechanisms. For the SILEX system planned by ESA the tracking sensor is required to determine the la position of the centre of the focused laser spot to within 0.1 μm on the CCD detector (equivalent to a mispointing of 0.07 μrad), for the nominal bandwidth of 8 kHz and the minimum optical power of 110 pW. This implies the determination of the centre position to small sub-pixel accuracies, achieved via mathematical interpolation from the fractions of the optical signal falling in the centre 4 pixels. This paper discusses the development of the tracking sensor breadboards by BAe and Sira. The detector unit, which uses the new 14x14 pixel CCD developed by Thomson-CSF, provides for fine adjustments of the CCD position. It also houses the electronics for CCD drive signals, output buffering, and amplification. The electronics unit carries out the processing of the video signal, to determine the coarse position of the spot over the 14x14 pixel area, and a very accurate position over the central 2x2 pixel area via an algorithm executed in a microprocessor. The paper also presents some performance results, and routes to the design of flight sensors.

One important subsystem for a transceiver package useable for optical space communication is the so-called Pointing, Acquisition and Tracking (PAT) Subsystem. Due to the highly collimated laser beams very precise beam steering is required. The PAT Subsystem is mainly involved in the acquisition procedure and has to guarantee the proper tracking of the line of sight so that the communication requirements, e.g. the bit error rate, can be met during the complete tracking phase. Furthermore the PAT Subsystem must be able to cover an angular range of more than a hemisphere if it is used on a low earth orbit (LEO) satellite which communicates with a geostationary (GEO) one. To meet the requirements of the PAT with respect to its dynamic and kinematic behaviour it is composed of three mechanisms which are the Coarse Pointing Assembly (CPA), the Fine Pointing Assembly (FPA) and the Point-Ahead Assembly (PAA). The interactions of these mechanisms are controlled by appropriate electronic and computing units which include the Control Law Electronics (CLE). This paper describes a complete PAT Subsystem for a CO2 laser communication package. All mechanisms and electronic units of this subsystem have been developed under ESTEC and DFVLR contracts. The most important features and test results are presented. Furthermore, design approaches of specific PAT mechanisms are illustrated. These mechanisms are developed within the scope of the SILEX project (Semi-conductor Laser Intersatellite Link Experiment) which is part of ESA's PSDE (Payload and Spacecraft Development and Experimentation) programme and aims at hardware which is different from the CO2 hardware.

The stringent pointing requirement for establishing an optical intersatellite link (ISL), leads to consider the dynamic interaction between host satellite and optical payload up to relatively high frequencies (f > 200 Hz ). For the optical pointing system design, this dynamic characterization is needed, and except measurements on Landsat 4 spacecraft, very few data are yet available. Cnes and Aerospatiale were associated in the frame of an Intelsat study to identify the dynamic environment of an optical payload on a geo-satellite. This paper presents a part of this study concerning the vibrations (high frequency, small amplitude) identification on a satellite, which is of major importance for the choice of the fine pointing loop rejection bandwidth. Testing and modelling were carried out together on TDFI spacecraft at Aerospatiale/Cannes facilities. Satellite's main body without appendages was mounted on a decoupling device, and high resolution sensors were fixed at potential ISL payload locations to record the disturbances induced by each vibrating source. To validate these test results, a dynamic modelling was performed. The spacecraft equations were generated in a. parametric form, thanks to the use of the symbolic computation software Gemmes. Effects of units which cannot be tested on ground (thrusters,...) are derived from this model. Comparison of tests results and computation are presented; both approaches appear very complementary. Correlations of both results allows to establish the disturbances distribution at the interface in term of amplitude and frequency.

The critical sub-microradian pointing accuracy required for an optical intersatellite link (ISL) at 0.8 μm can be resolved by a high bandwidth pointing loop inside the telescope. This loop, consisting of two high-speed fine sterring mirrors, must ensure either a short-time step response for large angular range (20ð peak to peak) in acquisition, or it must ensure accuracy and disturbance compensation in tracking. To verify the requirements, CNES developed in 1986 a breadboard representative of an actual european ISL, built and tested at ONERA. This paper describes the breadboard, the test bench and the experimental results for acquisition and tracking. A pointing loop was assembled from two General Scanning galvanometers and a charge coupled device matrix (CCD 7851 Thomson) to perform rallying, detection and ecartometry. A laser diode illuminates the CCD array and a He-Ne laser, lighting a quadrant cell , monitors the pointing error. The rallying and tracking phases, under ISL representative illumination were tested. Furthermore, pointing loop feasibility based on CCD detection was demonstrated. Computation results, measured in step response time, accuracy, and rejection, were verified using the breadboard. Since the system was subject to few data or input disturbances, a high 800 Hz bandwidth was obtained. Finally, bandwidth bottlenecks were clearly identified

Single quantum well laser diode arrays with emitting apertures of 200 um have operated up to 8 W cw. Monolithic multi-element arrays with emitting apertures of 1 cm are capable of cw operation up to 76 W cw. Devices fabricated using MOCVD epitaxial growth techniques and highly efficient single quantum well active region structures have exhibited high cw power and high efficiency. Catastrophic power limits of 8 W cw have been demonstrated by a 20 emitter multistripe laser diode with emitting aperture of 200 um. Similar structures consisting of 10 emitters in a 100 um aperture have operated to 6 W cw. (Fig. 1)

Results on phase-locked arrays of InGaAsP/InP index-guided diode lasers grown entirely by low pressure metalorganic chemical vapour deposition are presented. Two different array structures are compared : 1) the evanescently coupled array of inverted channeled substrate planar lasers and 2) the Y-junction coupled array of buried ridge lasers in order to assess their potential as sources of high powers in stable and narrow beam. Single-lobe beam operation is achieved by evanescent coupling, while the device is shown to be sensitive to fabrication and operation conditions and not practical as high power source. Uncoated Y-coupled arrays exhibit output power up to 158 mW CW/facet when bonded on diamond heatsink. Such devices are expected to give enhanced performance by further improvement in design and fabrication and should prove useful for free space optical communications.

In the frame of ESA's SILEX experiment a free space optical communication link based on 0.8 μm A1GaAs-laser diodes is planned for a maximum range of 45000 km and bit rates up to 65 Mbit/s. 30 mW average power is required from a single diode under QPPM modulation. Stable operation in the fundamental mode up to 120 mW pulse amplitude must be guaranteed and spectral width under modulation must not exceed 4 nm. Potential laser diodes had to be found, capable of meeting SILEX operating conditions. More than 15 laser manufacturers in Europe, USA, Canada and Japan have been contacted and almost 20 different diode types (single stripe) have been investigated and submitted to comprehensive measurements of Po/I-curve pulsed up to 120 mW, far field pattern, spectral spread under QPPM-modulation, mode hopping, astigmatism and rms wavefront error (WFE). Most of the laser diodes specified for 30 ... 50 mW cw output power have been delivered in a sealed TO 5-style package. The investigations showed that WFE is strongly affected by the window in introducing strong spherical aberration. Several diode types can be operated up to 120 mW peak power without kinks in the P,/I curve and meet the basic functional specifications set forth for SILEX. Based on the investigation results three laser types have been chosen for breadboarding and for accelerated life tests to overcome the lack of information on the lasers reliability under SILEX operating conditions.

Space optical communication requires high quality collimated laser diodes. A Hartmann-Shack wavefront sensor was developed at ONERA in order to test wavefront quality of such laser beams. For this application, the Hartmann-Shack wavefront sensor offers many advantages compared to an interferometer system (simple design, compactness, no moving part). Presently, the wavefront measurement precision is limited by the quality of the reference wavefront used to calibrate the Hartmann-Shack wavefront sensor (λ/100 rms). Preliminary experimental tests was achieved. The sensitivity of the Hartmann-Shack wavefront sensor was only limited by the experiment conditions.

A space-qualifiable broadband electrooptic phase modulator for CO2 laser intersatellite communication (wavelength 10.6 µm) was developed under ESA-contract. Eight Cadmium-Telluride crystals measuring 30 x 0.6 x 0.6 mm3 each were cascaded to yield a total active length of 240 mm. These crystals were embedded in a beryllia structure, designed to combine good heat transfer, high mechanical accuracy, and excellent electrical properties. Measured modulator characteristics are: optical transmittance > 80%, electrical bandwidth 2 GHz, maximum VSWR 1.75:1, modulation efficiency (at 20 W RF drive power) 1/8, modulation frequency response ±1 dB up to 1 GHz.

Silicon APD's (Avalanche Photodiodes) with very low excess noise factors have been developed at RCA. At 830 nm and an average gain of 150, an excess noise factor of 2.6 should be possible in a 100 μm active diameter detector, with a quantum efficiency of 90% and a capacitance of 0.25 pF. When combined with a low-noise preamplifier, it is predicted that direct-detection optical receivers can be constructed with a sensitivity which is within 3 to 4 dB of that achievable with an "ideal" optical receiver.

A 0.8 μm direct detection receiver front end (RFE) for operation in a system test bed (STB) for ESA's SILEX programme is presented. In the RFE a low capacitance high sensitivity silicon avalanche photo diode (APD) converts the weak optical receive signal (down to - 64 dBm average at 15 Mbit/s) to an electrical current. The APD is operated at a reverse bias voltage of 198 V yielding a multiplication factor around M=100 for maximum overall SNR. A low noise transimpedance (LNA) amplifier with GaAs MESFET input stage further amplifies this current. The feedback resistor is switchable with GaAs MESFETs for adaption of the overall response to bit rate and for optimization of sensitivity at the various bit rates. The LNA delivers a 10 mVpp range signal to a gain controlled main amplifier (MA) with a constant 0.5 Vpp output. The MA core is a custom designed monolithic integrated silicon variable gain broadband amplifier with a gain from 0 to 46 dB and a bandwidth from 100 kHz to 500 MHz. Both LNA and MA are realized using unpackaged chips mounted on thin film substrates in order to minimize stray and package capacitances and to provide good reproducibility. They are housed in a common COVAR package prepared for hermetical sealing. Gain control, bandwidth switching and switching-mode high-voltage APD bias supply are realized with specific circuitry and implemented on two separate PCBs contained in separate chambers of an aluminium package. The RFE consumes around 1.3 W of power, measures 51x70x75 mm3 and weighs 280 g. Mass and volume will be further reduced in the flight model design.

The 4PPM modulator/demodulator was developed as a breadboard model for the SILEX project of the ESA. The most important requirements placed upon this equipment is that it has to operate with any of four bit rates between 15 and 120 Mbit/s with minimal switch-over facilities, and that the demodulator has to accept from the optical receiver front end signals that are considerably distorted by band limitation. While the 4PPM modulator consists of a rather simple digital circuit, operable at all four bit rates, the 4PPM demodulator is composed of four units: The Input Section with bit rate-specific switchable filters, the Flash Comparator, where the selected input signal is digitized and sampled, the Maximum Likelihood Detector by which the maximal amplitude time slot is selected, and where the 4PPM/NRZ transformation is performed, and the Word-Error Integrator where the word clock is adjusted to its proper phase.

High-capacity intersatellite communication crosslinks will allow more efficient and reliable operation of military and commercial satellite systems. High-speed optical crossl inks can serve as a key building element of an interconnected space-based communication systemfor military applications. A network such as this would provide immediate communication among satellites, eliminating the need for ground-based relay stations and expensive worldwide ground tracking networks, which would greatly improve the efficacy and reduce the vulnerability of existing satellite systems. Crosslinks can also provide connectivty for commercial global satellite communication systems and for deep-space applications. Optical heterodyne communicatiol systems using semiconductor lasers offer small-aperture, modest-weight, low-power, point-to-point crosslink packages, characteristics that are suitable for the envisioned applications. System research and development performed at Lincoln Laboratory permits the implementation of an efficient optical crosslink based on readily available, state-of-the-art devices and technology.

This tutorial paper on coherent optical receivers first describes the principles of heterodyning, homodyning, and phase diversity reception. The limits for receiver sensitivity are stated for various binary modulation formats (ASK, FSK, PSK, DPSK) and their dependence on the linewidth-to-bit-rate ratio is illustrated. The requirements on the local laser oscillator are discussed, including the power necessary to make thermal noise negligible. The principle of the balanced photodetector is explained, and the question of heterodyne efficiency is addressed. Lastly the use of coherent detection for spatial acquisition and tracking is touched upon, and the advantage of coherent reception in the presence of background illumination is pointed out.

The theory and practice of free space optical mixing as applied to coherent intersatellite link receivers is addressed in this paper. Factors which contribute to mixing losses in free space are discussed, and theoretical predictions made. An experiment to measure the losses in a 1.5μm DFB coherent system is described, and preliminary results presented. Consideration is given to additional loss and distortion mechanisms which are not modelled, but which come into play for large off-axis angles.

Planetary and other deep-space missions of the future will require data return rates that far exceed current communication capabilities. Optical communications technology is expected to satisfy many of these needs. The overall architectural concepts and deployment strategies for deep-space optical communications is described. Design and development activities for the planetary spacecraft subsystem, as well as the Earth-vicinity reception system are then presented, along with some of the key supporting technology developments. Finally, the planning for deep-space optical communications systems developments and experimental demonstrations is discussed, as well as several of the technological challenges that must be overcome to fully realize the capabilities of this important technology.

The performance analysis of optical spaceborne communication systems shows that the main advantages of Nd:host-laser based IOL/ISL's resides in the high, scalable output power combined with a very narrow laser linewidth, which permits the implementation of very sensitive, coherent optical receivers. As laser output power will increase, the overall transmitter-laser and modulator-efficiency will gain importance and could finally decide the competitivity of the Nd:host laser technology. An overall laser reliability higher than that of the individual pumping diodes, as well as low transmitter mass and volume are additional important aspects. Preliminary experiments performed to assess the feasibility of both heterodyne and homodyne optical receivers were successfully carried out. A heterodyne optical receiver achieved shot noise limited operation and approached - after the IF-stage - the theoretical limit to AAdthin 4 dB to 5 dB. A simple optical PLL was also demonstrated. The residual phase errors were as low as 15° peak-to-peak with a filter loop bandwidth of 20 kHz. These laboratory results are part of a program which finally should lead to a representative and cost-efficient far field demonstration experiment for free-space Nd:Host laser communication systems.

Laser properties of Nd:LiNb03 were investigated and low loss proton exchange planar optical waveguides were fabricated in this material. Nd3+ spectroscopic properties were found to be unchanged indicating that the realization of a diode pumped Nd:LiNb03 waveguide laser should be possible.