Nicholas D. Hardy

Computational ghost imaging versus imaging laser radar for three-dimensional imaging

Ghost imaging has been receiving increasing interest for possible use as a remote-sensing system. There has been little comparison, however, between ghost imaging and the imaging laser radars with which it would be competing. Toward that end, this paper presents a performance comparison between a pulsed, computational ghost imager and a pulsed, floodlight-illumination imaging laser radar. Both are considered for range-resolving (3D) imaging of a collection of rough-surfaced objects at standoff ranges in the presence of atmospheric turbulence. Their spatial resolutions and signal-to-noise ratios are evaluated as functions of the system parameters, and these results are used to assess each systems performance trade-offs. Scenarios in which a reflective ghost-imaging system has advantages over a laser radar are identified.

Ghost imaging in reflection: resolution, contrast, and signal-to-noise ratio

Ghost imaging is a transverse imaging technique that relies on the correlation between a pair of light fields, one that has interacted with the object to be imaged and one that has not. Most ghost imaging experiments have been performed in transmission, and virtually all ghost imaging theory has addressed the transmissive case. Yet stand-off sensing applications require that the object be imaged in reflection. We use Gaussian-state analysis to develop expressions for the spatial resolution, image contrast, and signal-to-noise ratio for reflective ghost imaging with a pseudothermal light source and a rough-surfaced object that creates target-returns with fullydeveloped speckle. We compare our results to the corresponding behavior seen in transmissive ghost imaging, and we develop performance results for the reflective form of computational ghost imaging. We also provide a preliminary stand-off sensing performance comparison between reflective ghost imaging and a conventional direct-detection laser radar.

Classical far-field phase-sensitive ghost imaging.

We report the first (to our knowledge) far-field ghost images formed with phase-sensitive classical-state light and compare them with ghost images of the same object formed with conventional phase-insensitive classical-state light. To generate signal and reference beams with phase-sensitive cross correlation, we used a pair of synchronized spatial light modulators that imposed random, spatially varying, anticorrelated phase modulation on the outputs from 50-50 beam splitting of a laser beam. In agreement with theory, we found the phase-sensitive image to be inverted, whereas the phase-insensitive image is erect, with both having comparable spatial resolutions and signal-to-noise ratios.

A narrow-beam undersea laser communications field demonstration

We report a demonstration of narrow-beam laser communication through the waters of Narragansett Bay in Rhode Island, USA. The transmitter and receiver were mounted on an aluminum truss and placed in the water alongside a pier operated by the Naval Undersea Warfare Center. The transmitter consisted of a real-time modulator and encoder, a 515 nm wavelength commercial laser, collimating optics, and a steering mirror. The receiver included a steering mirror, a focal plane camera, a linear-mode avalanche photo-diode (APD), a photo-multiplier tube (PMT) single photon detector, a large area imaging camera, an iris to vary the field of view, optics to split the beam between the various detectors, and field-programmable gate array (FPGA) electronics for real-time demodulation and decoding. The PMT and APD detectors were used for communications demonstrations; the imaging and focal plane cameras were used for channel characterization measurements and system alignment. Communications and characterization data were collected through a variety of conditions over the five day field experiment, including day and night, calm and high winds, and flood and ebb tide. In the experiment, the transmit power, receiver field of view, and link distance were varied. The water transmissivity and volume scattering function were measured throughout the experiment to calibrate the results. Real-time communications demonstrations with the PMT were carried out between 1 megabit-per-second (Mbps) and 8.7 Mbps at 7.8 meters, which represented between 8 and 12 beam extinction lengths. With the APD, 125 Mbps were demonstrated at 4.8 meters, representing approximately 5 extinction lengths.

Turbid-harbor demonstration of transceiver technologies for wide dynamic range undersea laser communications

Undersea laser communications represent a promising area of research with a large set of applications. Wide dynamic range receivers are necessary to operate through a range of possible water qualities and link distances. In the signal-starved regime, photon-counting photomultiplier tubes (PMTs) are a key technology for high-sensitivity communications. When more signal is available, linear avalanche photodiodes (APDs) provide an opportunity for higher-rate communication. We have designed a receiver terminal employing both kinds of detectors to show robust operation over nearly two orders of magnitude in power and data rate. An optical link including this receiver terminal was submerged in Narragansett Bay, RI to demonstrate underwater optical communication over several days. The PMT receiver demonstrated robust, error-free performance over channel rates from 1.302 Mbaud to 10.416 Mbaud for received optical power levels ranging from -84.1 dBm to -75.3 dBm. The PMT link demonstrated an error-free user rate of 8.68 Mb/s. This corresponded to nearly-ideal detector efficiency on the order of one detected photon per bit. The PMT receiver was contained entirely within the submerged enclosure and demonstrated full real-time decoding, including strong forward error correction. A low-power transmitter was used to demonstrate a link with loss equivalent to 18 extinction lengths. With moderately-powered transmitters, this distance could be extended to 22.4 extinction lengths. The PMT receiver was capable of operating at near-theoretical limits during the day and night. Its multi-rate operation demonstrated the capability of trading sensitivity for data rate efficiently. With the same low-power transmitter, the APD receiver achieved a bit error rate less than 1×10-9 at 125 Mbaud. Furthermore, it achieved an error rate correctable by forward error correction for a link with loss equivalent to 9 extinction lengths.

Modeling and experimental validation of narrow beam propagation through a turbid harbor

Narrow-beam laser communication (lasercom) can enable high-rate, long-range undersea communication. Light in the beam is absorbed and scattered by suspended particulates, altering the spatial, angular, and temporal properties of the transmitted light. Development of high quality data links is aided by accurate simulations that can be adapted to different water conditions, from turbid harbors, to clear oceans, and even laboratory test beds with artificial scatterers. To this end we have developed a method to create an empirical scattering function from volume scattering function measurements that can be used to generate random scattering angles in an undersea optical propagation simulator. This method is presented, along with cross validation of its predictions during a recent narrow-beam communication demonstration in Narragansett Bay, Middletown, RI, USA. VSF measurements taken during the trials were used to run the simulator and the results are shown to agree with in situ images of the pupil and focal planes of the communication laser beam.

A burst-mode photon counting receiver with automatic channel estimation and bit rate detection

We demonstrate a multi-rate burst-mode photon-counting receiver for undersea communication at data rates up to 10.416 Mb/s over a 30-foot water channel. To the best of our knowledge, this is the first demonstration of burst-mode photon-counting communication. With added attenuation, the maximum link loss is 97.1 dB at λ=517 nm. In clear ocean water, this equates to link distances up to 148 meters. For λ=470 nm, the achievable link distance in clear ocean water is 450 meters. The receiver incorporates soft-decision forward error correction (FEC) based on a product code of an inner LDPC code and an outer BCH code. The FEC supports multiple code rates to achieve error-free performance. We have selected a burst-mode receiver architecture to provide robust performance with respect to unpredictable channel obstructions. The receiver is capable of on-the-fly data rate detection and adapts to changing levels of signal and background light. The receiver updates its phase alignment and channel estimates every 1.6 ms, allowing for rapid changes in water quality as well as motion between transmitter and receiver. We demonstrate on-the-fly rate detection, channel BER within 0.2 dB of theory across all data rates, and error-free performance within 1.82 dB of soft-decision capacity across all tested code rates. All signal processing is done in FPGAs and runs continuously in real time.

Propagation modeling results for narrow-beam undersea laser communications

Communication links through ocean waters are challenging due to undersea propagation physics. Undersea optical communications at blue or green wavelengths can achieve high data rates (megabit- to gigabit-per-second class links) despite the challenging undersea medium. Absorption and scattering in ocean waters attenuate optical signals and distort the waveform through dense multipath. The exponential propagation loss and the temporal spread due to multipath limit the achievable link distance and data rate. In this paper, we describe the Monte Carlo modeling of the undersea scattering and absorption channel. We model photon signal attenuation levels, spatial photon distributions, time of arrival statistics, and angle of arrival statistics for a variety of lasercom scenarios through both clear and turbid water environments. Modeling results inform the design options for an undersea optical communication system, particularly illustrating the advantages of narrow-beam lasers compared to wide beam methods (e.g. LED sources). The modeled pupil plane and focal plane photon arrival distributions enable beam tracking techniques for robust pointing solutions, even in highly scattering harbor waters. Laser communication with collimated beams maximizes the photon transfer through the scattering medium and enables spatial and temporal filters to minimize waveform distortion and background interference.

Sub-wavelength stabilization of long, deployed optical fibers for quantum networks

We implemented an active feedback scheme to stabilize an ∼84 km deployed optical fiber between Lincoln Laboratory and MIT Campus. The residual fluctuations of less than 193 attoseconds RMS enable quantum networking and quantum secure communications.

Undersea narrow-beam optical communications field demonstration

Optical propagation through the ocean encounters significant absorption and scattering; the impact is exponential signal attenuation and temporal broadening, limiting the maximum link range and the achievable data rate, respectively. MIT Lincoln Laboratory is developing narrow-beam lasercom for the undersea environment, where a collimated transmit beam is precisely pointed to the receive terminal. This approach directly contrasts with the more commonly demonstrated approach, where the transmit light is sent over a wide angle, avoiding precise pointing requirements but reducing the achievable range and data rate. Two advantages of narrow-beam lasercom are the maximization of light collected at the receiver and the ability to mitigate the impact of background light by spatial filtering. Precision pointing will be accomplished by bi-directional transmission and tracking loops on each terminal, a methodology used to great effect in atmospheric and space lasercom systems. By solving the pointing and tracking problem, we can extend the link range and increase the data throughput.