The Deep Space Optical Communication (DSOC) project will demonstrate free-space optical communication at almost 3 AU, or 3 orders of magnitude further than any previous attempt. DSOC will utilize the 5m Palomar Hale Telescope to receive the downlink signal, which will couple the downlink light onto an optical table and into a superconducting nanowire single photon detector (SNSPD). The output of the SNSPD is digitized by the Ground Laser Receiver Signal Processing Assembly (GSPA) using a high throughput streaming time to digital converter (TDC). The GSPA is a scalable FPGA-based receiver which demodulates and decodes the DSOC downlink signal through novel signal processing algorithms implemented on Xilinx UltraScale+ FPGAs, as well as Python-based software monitor and control routines. Exploiting the unique TDC-based architecture, the GSPA supports over four orders of magnitude of downlink data rates across multiple orders of magnitude of signal and background powers. In this paper we present an overview of the hardware, firmware and software architectures to implement this system, as well as performance analysis for links ranging from near-Earth to 2.8 AU.

This paper discusses recent activities at JPL that are focused on extending the Opportunistic Multiple Spacecraft Per Antenna (OMSPA) concept to include arraying multiple antennas. Specifically, we explore the ability to process multiple open loop recordings associated with multiple antennas and perform the appropriate alignment and combining. We focus on using the symbol stream combining technique and provide examples of performance measurements on actual spacecraft signals for MarCO A and B as well as the Mars Express.

As smallsats become increasingly capable, longer-lived, and have more secondary payload launch opportunities to beyond-GEO destinations, they are expected to play an increasing role in deep space science investigations. This expectation is borne out by several relatively recent NASA Science Mission Directorate solicitations regarding smallsat studies and small innovative missions. With the potential for these smallsats to substantially add to the number of spacecraft operating in deep space, we need to be thinking about ways to support communications with all of them without the huge expense of trying to build a commensurate number of deep space antennas. One approach to this challenge might involve making greater use of beam-sharing techniques that allow all the spacecraft within the beamwidth of a single ground antenna to simultaneously downlink to the antenna. One of these techniques, Opportunistic Multiple Spacecraft Per Antenna (OMSPA), may be particularly suited to smallsats. In the concept for this technique, smallsats within the scheduled ground antenna beam of some other spacecraft, make opportunistic use of that spacecraft's beam by transmitting "open-loop" to a recorder associated with the antenna. These transmissions get captured on the recorder and can be later retrieved, demodulated, and decoded so that the smallsats can recover their data - all without them having to schedule the antenna itself and compete with larger missions for antenna time. Widespread use of such a technique could lead to more efficient use of receiver antenna resources and result in a dramatic increase in downlink throughput. An opportunity to demonstrate the technique occurred in May 2018, when the Mars CubeSat One (MarCO) mission, consisting of two nanospacecraft (MarCO-A & B) launched alongside InSight, a NASA Mars lander mission. To demonstrate the capabilities of OMSPA for this launch window opportunity, X-band downlink telemetry was recorded for all three spacecraft (InSight, MarCO-A, and MarCO-B) at both the Deep Space Network (DSN), using its 34-m antennas, and at Morehead State University (MSU) using its 21-m antenna - with all of the involved antennas pointed at InSight. Open-loop recordings were collected at the DSN using wideband very long baseline science receivers and at MSU using Universal Software Radio Peripheral (USRP) devices operated using GNU Radio. All the recordings were then processed at the Jet Propulsion Laboratory, California Institute of Technology (JPL) using an OMSPA Software Receiver, a signal processing/communications tool used to extract telemetry transfer frames from baseband samples. The results of extracting telemetry data from InSight/MarCO recordings collected by the DSN and at MSU are described in this article. In particular, details pertaining to the processing chain used by the OMSPA Software Receiver to demodulate the DSN and MSU recordings are presented, from carrier/symbol synchronization, to frame alignment using attached sync markers (ASMs), followed by error correction code decoding. Validation results with closed-loop data obtained by the DSN are also presented in order to highlight the viability of OMSPA for future multiple spacecraft demodulation opportunities.

— The purpose of this article is to summarize the attempts made to carry out spectrum reconstruction for interferograms obtained from the Tropospheric Emission Spectrometer (TES) on-board the Aura satellite, after the laser used to trigger the sampling unit malfunctioned in the spring of 2016. In particular, we highlight the drawbacks that made reconstruction problematic and the approaches attempted to overcome them. These drawbacks include ill-conditioning of the reconstruction problem incurred with nonuniformly spaced interferogram samples, discrepancies in the reported and true spatial locations of the samples, and low resolution of the spatial location data. Algorithms used to overcome these pitfalls, involving regularized reconstruction methods in the case of unknown/imperfect sample locations, are presented, along with reconstruction results showing the difficulties associated with the spectrum reconstruction problem.

— This article reports on the development and validation of variable coded modulation (VCM) on the Jet Propulsion Laboratory (JPL) software-defined radio (SDR). Flight tests were conducted to evaluate performance of the VCM waveform over an S-band link between the Space Communications and Navigation (SCaN) Testbed and the Glenn Research Center (GRC). The tests verified the VCM waveform’s ability to switch to different modulation and coding modes adapting to varying link conditions, and demonstrated improvement in effective data throughput as compared to NASA standard waveforms. We also describe a suite of ground receiver tools that were developed to autonomously acquire, track, and post-process the JPL VCM waveform. Processing results from one of the tests indicate an overall improvement of ∼ 2 dB in data throughput over standard waveforms. The demonstrated technologies are building blocks of a future cognitive radio system with the capability to adapt its operation to the communication environment in near real time.

The goal of the Deep Space Optical Communications project at the Jet Propulsion Laboratory is to demonstrate laser communication links at ranges out to approximately 3 AU. In this paper, we discuss a downlink receiver concept capable of demodulating optical pulse-position modulated (PPM) waveforms with data rates varying from approximately 50 kbps up to 265 Mbps, using a range of PPM orders, slot widths, and code rates. The receiver operates on recorded timestamps corresponding to the times-of-arrival of photons detected by a photon-counting detector array followed by a commercial time-tagger. Algorithms are presented for slot, symbol, and frame synchronization as well as parameter estimation. Estimates of link performance are evaluated through Monte Carlo simulation for an optical channel that includes optical losses, detector blocking, signal clock dynamics, and pointing-induced downlink fades. Based upon these simulation results, it is expected that link closure may be achieved with at least 3 dB of margin under a variety of relevant conditions.

A compact silicon photonic channelized optical spectrum monitor is designed and realized, which can replace a large rack-mounted OSA's channel power monitoring functionality, and the signal processing algorithm underlying its operation is described.

Summary fom only given. Recent passive space-borne microwave observing systems operating below 40GHz have shown an increase in the amount of man-made interference corrupting incoming natural thermal emissions (McKague et al., 2010 IGARSS). Many radiometer systems operate in bands (e.g. 18.7GHz) that are shared with space to ground-transmissions. Other space-borne systems (e.g. Aquarius, Soil Moisture Ocean Salinity - SMOS) operate in protected radio bands to avoid Radio Frequency Interference (RFI). Measurements from these missions have shown that RFI still persists even in protected bands. The RFI environment has forced many radiometer systems to operate in narrower bands than usual. This directly impacts the radiometric noise and instrument design, which in turn impacts the necessary fidelity required for retrieving the EDRs. Based on these issues, there is a need for developing wideband microwave radiometer systems that can co-exist with a harsh RFI environment. The following talk will present the work undertaken by the Jet Propulsion Laboratory, to develop an agile wideband digital backend system that can operate in and adapt to any RFI environment. The digital backend system needs to be capable of implementing a flexible digital signal processing system that can detect and mitigate RFI contaminated spectrum regions. The goal of the digital backend is to incorporate all necessary processing in the backend to take a corrupted spectrum and produce a single RFI mitigated output value with a minimal data rate. The first portion of the talk will focus on the intermediate RFI detection algorithms that were compared and contrasted with each other in terms of algorithm performance and backend implementability. The algorithms are compared with respect to various RFI parameters such as duty-cycle, power, spectral width, number of sources etc. We utilize innovative evaluation techniques and performance metrics to compare the different algorithms. The algorithms are also tested using real airborne data measured during the Soil Moisture Active/Passive Validation Experiment (SMAPVEX) field campaign of 2012. An optimal version of the kurtosis detection algorithm and an innovative “squrtosis” algorithm with cross-frequency is implemented. A brief description of these algorithms will be presented. The final aspect of the talk will focus on the firmware implementation of the above algorithms. The algorithms are implemented on a Reconfigurable Open Architecture Computing Hardware (ROACH) -2, a Xilinx Virtex 6 stand-alone FPGA board. We will present results on the initial implementation as well as initial results based on lab-generated RFI signals. Further work based on the obtained results will also be discussed.

In this article, we develop and analyze an uplink signal detection technique for the Deep-Space Optical Transceiver (DOT). Here, the detection is carried out using a set of test statistics obtained from up-down counter (UDC) photon detection systems. Specifically, we address two sets of statistics: the count outputs from a bank of uniformly temporally spaced UDCs as well as the counts from a single UDC that cycles through multiple uniformly spaced timing phases. From these test statistics, we derive the Neyman-Pearson decision rule under certain input conditions and analyze the performance of this hypothesis test. We show the performance trade-offs associated with both sets of test statistics, which can then be used to determine which set to use as well as the number of UDCs or timing phases required for implementation.

In this paper, we present a complex baseband model for a wideband power amplifier that incorporates carrier frequency dependent amplitude modulation (AM) and phase modulation (PM) (i.e., AM/AM and AM/PM) characteristics in the design process. The structure used to implement the amplifier model is a Wiener system which accounts for memory effects caused by the frequency selective nature of the amplifier, in addition to the nonlinearities caused by gain compression and saturation. By utilizing piecewise polynomial nonlinearities in the structure, it is shown how to construct the Wiener model to exactly accommodate all given AM/AM and AM/PM measurement constraints. Simulation results using data from a 50 W 32-way Ka-band solid state power amplifier (SSPA) are provided, highlighting the differences in degradation incurred for a wideband input signal as compared with a narrowband input.

The research described in this publication was carried out by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. © 2012 California Institute of Technology. U.S. Government sponsorship acknowledged. abstract. — The Reconfigurable Wideband Ground Receiver (RWGR) system is a variabledata-rate, reprogrammable hardware/software receiver developed to meet the wide range of telemetry demodulation processing needs of the Deep Space Network (DSN). In this article, we report results from a number of field tests with the RWGR deployed to the Deep Space Station 13 (DSS-13) experimental ground station.

Deep-space navigation uses estimates of range and Doppler to update and improve spacecraft trajectory solutions. However, the transmission of tones or PN sequences drain power and bandwidth that could be better used for transmitting additional science data from the spacecraft. Our scheme uses a conventional uplink ranging signal, but the downlink is replaced with an asynchronous telemetry signal whose timing relative to the acquired uplink signal is measured. This measurement, along with the acquired timing of the received telemetry, enables the round-trip light-time to be computed on the ground. In this paper, the structure of the joint maximum likelihood estimator for range and Doppler is derived, and its performance determined relative to Cramer-Rao bounds via simulation and analysis. Performance of individual frequency estimators based on conventional Costas loop phase estimates where the delay is assumed to be known, and of delay-tracking loops that assume known frequency and phase are also derived, and contrasted with the performance of the optimum Doppler-delay estimator. Advantages of this new approach include the ability to simultaneously collect ranging measurements and transmit the highest supported telemetry rate throughout the pass, and compatibility with suppressed carrier signaling and higher order modulations. This approach could result in significant additional ranging data and total data volume return for future missions.

Deep-space navigation uses estimates of range and Doppler to update and improve spacecraft trajectory solutions. (1,2) Operationally, Doppler is generally extracted directly from the ground receiver's carrier tracking loop, and range is determined primarily by the use of specially designed "ranging tones," or more recently (e.g., on New Horizons), Pseudo-Noise (PN) sequences. Transmission of tones or PN sequences drain power and bandwidth that could be better used for transmitting additional science data from the spacecraft. Here we describe and evaluate a novel technique that extracts range and Doppler estimates directly from the decoded data, thus enabling data-transmission at the maximum rate consistent with spacecraft range, antenna gain and available signal power. In this paper, the structure of the maximum likelihood estimator for range and Doppler is derived, and its performance determined relative to Cramer-Rao bounds via simulation and analysis. Performance of conventional Doppler and delay estimators based on carrier tracking loops is also derived, and contrasted with the performance of the optimum Doppler-delay estimator over a range of symbol SNR typically encountered in deep-space applications.

Recently, much effort has been placed toward the development of the Reconfigurable Wideband Ground Receiver (RWGR): a variable-rate, reprogrammable software-defined radio intended to supplement and augment the capabilities of the Block V Receiver. In this report, we first give an overview of the hardware architecture of the RWGR, including a detailed description of the filter-decimate front-end and subsequent high-rate receiver system, which includes a 4 sample/symbol demodulator core. We then present a series of laboratory hardware performance results, including bit error rate (BER) results at various data rates. The RWGR is shown to yield performance close to theory in terms of BER with losses typically less than 1 dB in bit signal-to-noise ratio (SNR).

In this paper, we derive joint parameter estimation bounds for the symbol timing offset, carrier phase, and carrier frequency offset for any shaped offset quadrature phase shift keying (SOQPSK) waveform. Specifically, we calculate the conditional Cramér-Rao bound (CCRB) for a known given data sequence and compare this to the modified Cramér-Rao bound (MCRB) for the case in which the data pattern is unknown and random. This allows us to assess the performance of candidate preamble waveforms for parameter acquisition. We show how to simplify the CCRB when the preamble waveform is periodic and compare the CCRB performance of several candidate preambles to the MCRB. For the Telemetry Group (TG) variant of SOQPSK (i.e., SOQPSK-TG), we specifically show that certain candidate preambles can offer substantial improvements over the MCRB in terms of symbol timing offset estimation with negligible adverse impact on the estimation of the carrier phase and carrier frequency offset.

In this article we characterize the effect of transmitter clock jitter upon receiver symbol synchronization performance. Using a sinusoidal model for the timing jitter, we evaluate the bit error rate (BER) degradation and cycle slip probabilities of receivers via analysis as well as simulation for uncoded offset quadrature-phase-shift-keying (OQPSK). We evaluate performance for two different symbol synchronization loops: the modified data transition tracking loop (M-DTTL) and the Gardner loop. The results are parameterized in terms of the timing jitter parameters (peak frequency jitter, time interval error, and cycle-to-cycle jitter) as well as symbol tracking loop parameters (loop damping factor, loop bandwidth). We present analytical expressions for BER degradation in the presence of sinusoidal timing jitter and compare results with those obtained via simulation, as well as past hardware tests of receivers. These results show that for both types of symbol synchronizers, peak BER degradation decreases as the loop damping factor increases, and that for underdamped tracking loops, the BER degradation peaks when the normalized jitter rate is approximately the same as the natural frequency of the loop transfer function. Simulated cycle-slip rates are also presented, showing the effects of varying loop bandwidths and damping factors. Finally, we illustrate how BER degradation can be characterized in terms of jitter time interval error and cycle-to-cycle jitter, providing predictive capabilities for receiver performance and guidelines for the specification of transmitter clock requirements.

The research described in this publication was carried out by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. © 2010 California Institute of Technology. Government sponsorship acknowledged. Recently, much effort has been placed toward the development of the Reconfigurable Wideband Ground Receiver (RWGR): a variable-data-rate, reprogrammable receiver, whose technologies are intended for infusion into the Deep Space Network. A significant thrust of that effort has been focused on the development of field-programmable gate array (FPGA)–based algorithms for processing high-rate waveforms up to 640 Mbps. In this article, we describe the development of software receiver algorithms used to perform telemetry demodulation of lowto medium-data-rate signals.

The eigenvalue decomposition (EVD) of a Hermitian matrix in terms of unitary matrices is well known. In this paper, we present an algorithm for the approximate EVD (AEVD) of a para-Hermitian (PH) system. Here, the approximate diagonalization is carried out successively by applying degree-1 finite impulse response (FIR) paraunitary (PU) transformations. The system parameters are chosen to make the zeroth order diagonal energy (ZODE) nondecreasing at each stage. Simulation results presented for the design of a signal-adapted PU filter bank (FB) show close agreement with the behavior of the infinite order principal component FB (PCFB).