A Prototype Optical SETI Observatory
Proceedings of SPIE's 1996 Symposium on Lasers and Integrated Optoelectronics, Photonics West '96, The Search for Extraterrestrial Intelligence (SETI) in the Optical Spectrum II, Vol. 2704, San Jose, California, January 27-February 2, 1996, pp. 102-116.
Copyright ©, 1996, The Columbus Optical SETI Observatory
Stuart A. Kingsley
The Columbus Optical SETI Observatory
The Optical Search for Extraterrestrial Intelligence (OSETI) is based on the premise that there are ETIs within our galaxy which are targeting star systems like our own with free-space laser beams. Upon these beams will ride attention-getting beacon signals and wideband data channels. Perhaps the wideband channels form part of a Galactic Information Superhighway - a Galactic Internet to which we are presently oblivious. The Columbus Optical SETI (COSETI) Observatory described in this paper is intended to be a prototype observatory which might lead to a new renaissance in both optical SETI and optical astronomy. It is hoped that the observatory design will be emulated by both the professional and amateur communities. The modern-day OSETI observatory is one that is more affordable than ever. With the aid of reasonably priced automatic telescopes, low-cost PCs, software and signal processing boards, Optical SETI can become accessible to all nations, professional scientific groups, amateur astronomy societies and even individuals.
Keywords: Optical, SETI, OSETI, astronomy, lasers, Columbus, observatory, LX200, SCT, CCD
In the summer of 1990, the author commenced revisiting the optical approach to the Search for Extraterrestrial Intelligence (Optical SETI), an idea almost as old as modern-day Microwave SETI. Star Trek has an impressive history of stretching the viewer's imagination with regard to the possibilities that exist out in deep space. The author has long believed in the existence of Extraterrestrial Intelligence (ETI), an idea that Star Trek's ingenuity has made seem almost obvious, and hopes that one day it will even become possible for mankind to travel between the stars - that we will not be locked forever within the confines of our star system. However, if it is not found to be possible to travel at warp speed, discover worm holes that might be used as short cuts through the back of the universe or develop so-called "sub-space communications", then electromagnetic communications will probably be the only means we have to establish contact with other intelligent civilizations.1. Nevertheless, compared with actual space travel, the development of laser technology by ETIs is trivial and should be non-taxing to their technical prowess. Free-space laser communication for them would be no more difficult or expensive than its microwave counterpart. At any rate, in an infinite universe, with infinite possibilities, it might well be that while some civilizations invent "warp-speed" technology or discover "worm-holes", others stay home and just communicate.
The age of Microwave SETI can be said to have started in 1959, while the still-born Optical SETI first saw the light of day in 1961. For various reasons, by the early '70's, SETI rationales became based on the microwave approach, and the viability of the visible or infrared laser approach had largely been discounted. A review of the SETI literature of the past 34 years appears to indicate that the report on Project Cyclops had much to do with the subsequent lack of interest in the optical approach to SETI. In those days, the word "optical" was still synonymous with the word "visible". However, over the past two decades, as the fiber-optics industry has matured, "optical" has been redefined to cover the entire electromagnetic regime from the far-infrared to the ultra-violet. The technology of "Photonics" or "Optoelectronics" is the major force that is shrinking our planet today, aka the "Global Village". It is in this redefined sense that the word "optical" is used here - a superset of both the visible and infrared regimes. Perhaps within the Milky Way, the Encyclopedia Galactica is being exchanged between civilizations on an optically-based Internet, but we have yet to find or establish a node!
For sure, the future of terrestrial communications is one that is largely photonics based. Some time in the next 30 years, men will for the first time walk upon the surface of Mars. The historic High Definition (HDTV) pictures of that event will be transmitted back to Earth via a laser link, probably from the Martian surface to a Martian orbiter via laser, from the orbiter to Earth via laser and around the Earth via laser-linked geostationary satellites and into peoples' home via fiber-optics. Truly, the future is photonic. Photonic for our terrestrially-based future and so too for our space-based future. To argue otherwise is to fly in the face of reason.
Initially, the author compared the effectiveness of free-space interstellar communications using visible and infrared lasers operating in a very monochromatic continuous wave (cw) mode and with the use of heterodyne detection techniques, with that of radio-frequency systems. By January 1993, after communication with Monte Ross, and SPIE's first international conference on Optical SETI, the author decided to concentrate on the assumption that most ETIs employed pulsed lasers and that our receiving systems should employ incoherent photon-counting techniques. It can easily be shown that pulsed lasers are very effective for both beacon signals and wideband channels. The main advantage is that the high peak to mean power levels, allow for the received pulses to easily overcome the effects of photon noise caused by the sky and stellar backgrounds, without the need to employ narrow-band optical filters. This means that there is no need to postulate a "magic frequency" or "magic wavelength" - only a "magic wavelength regime". It also obviates the need to employ complex and expensive coherent optical heterodyne techniques with local oscillator lasers.
Sometimes it is overlooked that most discussion of SETI involves the reception of "beacon signals", not the more wideband channel that would convey rich data. That there should be some connection between the "beacon" channel and the "data" channel is often ignored. It is the latter that will probably govern the placement of both signals in the electromagnetic spectrum. Indeed, without full consideration of how a wideband signal can be transmitted over interstellar distances, we are likely to draw erroneous conclusions about so-called "magic wavelengths or frequencies".
1.1 Some Philosophy
It is the author's present belief that as far as interstellar electromagnetic communications are concerned, only targeted free-space laser communications can sustain the signal-to-noise ratio, and the low interstellar dispersion and scintillation effects, so as to convey wideband, Gbit/s type data links over hundreds or thousands of light years. Using microwave transmitters to "broadcast" signals across interstellar space is very inefficient, and cannot support wideband communications due to insufficient signal-to-noise ratio, and significant interstellar dispersion and scintillation effects.
A negative outcome of a visible SETI search would not imply that ETIs do not exist or use lasers, but merely that their laser transmissions may be in the infrared or far-infrared, even at wavelengths to which the atmosphere is not transparent. It may well be that space-based receivers will be required for successful reception of SETI signals. The atmospheres of the intended targeted planets might be used as a safety blanket to delay successful reception of the signals until the targeted civilization was mature enough to have developed space-based technologies, and presumably, be in a better condition to avoid catastrophic effects from too early a cultural contamination.
In October 1992, coincident with the 500th anniversary of Christopher Columbus's discovery of the New World, and the switch-on of NASA's Microwave SETI Project (initially called MOP, then HRMS and then reborn as the privatized Project Phoenix), the author decided to commence the design, funding and construction of his own prototype Optical SETI (OSETI) observatory. Now, three years later, the basic construction of The Columbus Optical SETI (COSETI) Observatory is nearing completion. It is expected that the first data will be obtained in the spring of 1996. The idea is that the observatory will form the basis for a rekindling of interest in the optical approach to SETI within the professional and amateur astronomical community. This paper summarizes the rationale behind this effort (previously given in detail, elsewhere) and describes the basic design of the observatory. It should be noted that the design has been affected by the dual-track nature of this project - in that it must serve to generate interest in the professional scientific and engineering communities, and yet appeal to the amateur scientific and astronomy communities. To a large extent, the differences are substantially dictated by economics, i.e., professional systems are more prohibitively expensive than amateur systems. However, since at this time there are essentially no professional visible or near-infrared Optical SETI observatories, the amateur community is expected to make a major contribution to this field of investigation.2-7
1.2 The Optical SETI Rationale
It is assumed that extraterrestrial civilizations would combine an attention-getting laser beacon signal with a wideband data channel. Furthermore, it is thought reasonable to suggest that ETIs would have powerful nuclear or stellar-pumped laser transmitting systems in orbit about their stars.
For a pulse repetition rate of 1 Hz, these lasers could be capable of producing nanosecond duration pulses with peak beacon powers of the order of 1018 W, and with corresponding peak Effective Isotropic Radiated Powers (EIRPs) greater than 1033 W. During each pulse, such lasers could outshine the ETIs' star by seven orders of magnitude! Such powerful pulses could allow for signal beacon detection by small terrestrial telescopes of about 1,000 photons per pulse at a range of 100 light years. Thus, it is conjectured that small receivers, such as that based on the 10" (25.4 cm) aperture Meade LX200 Schmidt-Cassegrain Telescope (SCT) used at The Columbus Optical SETI (COSETI) Observatory, would be capable of detecting such beacons out to a range of about 1,000 light years. This range encompasses approximately one million solar-type stars.
This pioneering observatory will shortly begin examining the light emanating from several solar-type stars within 100 light years of the Sun. These stars, which appear in the official microwave targeted search list produced by the SETI Institute, are initially being analyzed for repetitive laser pulses in the 500 nm region of the visible spectrum. It is expected that much larger receiving telescope apertures would be required to successfully detect the associated wideband data channel.
In the prototype Optical SETI observatory, photons are detected by a photomultiplier and counted with the aid of a discriminator set at a high threshold level. Optical bandpass filtering may be employed to further reduce the background "noise" count. The high discriminator threshold and bandpass filtering are required in order to mitigate the vast storage requirements that would be necessary if all stellar and sky background photons were detected and counted. At the COSETI Observatory, telescope control and signal processing are presently performed using two Windows-based 486 PCs, EPOCH 2000 astronomy software, a GageScope CSLITE PC oscilloscope board and LabVIEW for Windows. Time-domain, frequency-domain and statistical analyses will be employed to examine the signals for the presence of repetitive pulses.
In this section we show how "easy" it is to send a strong pulsed laser beacon signal over a thousand light years and detect such a beacon with a small aperture telescope. Presently, the highest peak power that mankind can produce is of the order of 1012 W. It has taken about 40 years to go from a 3 kW ruby laser pulse to a 1012 W Nd:YAG pulse produced by the NOVA facility at Lawrence Livermore Laboratories. Early next century, the National Ignition Facility (NIF) at Lawrence Livermore would be capable of producing 1 ns type pulses of the order of 1015 W peak power at the rate of only one pulse per day. How much longer it will take us to develop the capability to produce pulses of 1018 W peak power at a repetition rate of 1 Hz is hard to estimate, but it is likely to occur sometime early next century - no time at all on the cosmic time scale!
Then, with other enabling technologies and capabilities, e.g., the existence of space-based optical telescope arrays allowing the direct imaging of the planetary systems around nearby stars, we too will have the ability to point-ahead target, and to aim and strike the planetary systems of stars within a thousand light years. We will then be in the age of CETI - Communications with Extraterrestrial Intelligence in other parts of our galaxy. All this may occur before our first space probe to Alpha Centauri sends back its "live" video and data of that star's planetary system.
The relationship peak power Ppk and average power for a pulsed laser is given by:
Pav Ppk = ------ W (1) tau.r
t = pulse width (1 ns),
Substituting the values in parentheses for a pulsed ETI laser beacon system with a 1 Hz repetition rate, we find that:
Ppk = 1018 W.
Diffraction-Limited Telescope Gain
The gain of a diffraction-limited dish or telescope is given by:
pi.D G = [----]2 (2) Wl
D = diameter of transmitter
aperture (10 m),
Substituting the values in parentheses for an ETI uplink transmitting at the center of the human photopic response, we find that:
G = 155.1 dB.
Note that it has been suggested by the late Barney Oliver that it would be impossible for ETIs to make use of such high antenna gains. Indeed, at the OSETI I conference4, Barney Oliver indicated that the upper bound for usable uplink gain, be it for a microwave or optical system, would be about 74 dB. This was subsequently increased to 94 dB in light of recent improvements in our knowledge of the peculiar proper motions of nearby stars. This author strongly disagrees with such low limits on antenna gain. Indeed, the arbitrary gain of 155 dB used by the author may not mark the maximum usable gains that ETIs can employ, either with a single uplink telescope or a phased-array of telescope elements. See the OSETI I proceedings for details about Barney's arguments and the author's counter-arguments.
Effective Isotropic Radiated Power
The Effective Isotropic Radiated Power is the power that the transmitter appears to have if it radiated isotropically. It is given by:
EIRP = P.G W (3)
Substituting the values for P and G
given above, we find that:
EIRPlaser = 3.2 x 1033 W.
Note that for a star like the sun:
EIRPstar = 3.9 x 1026 W.
So we note that during each brief 1ns pulse, the ETI laser outshines its star by about seven orders of magnitude. Thus, in principle, there is no problem in discerning the signal pulse from the "noise" of the stellar background.
The intensity of the received signal and stellar background noise is given by:
EIRP I = -------- W/m2 (4) 4.pi.R2
where R = range (9.461 x 1016 m).
Substituting the values in parentheses for a range of 10 light years, we find that just outside the atmosphere:
Ilaser = 2.8 x 10-2 W/m2.
Istar = 3.5 x 10-9 W/m2.
Detected Signal Power
The signal power appearing at the photodetector through a V-type optical filter is given by:
Tatm = atmospheric
Substituting the values in
parentheses, we find that:
Slaser = 8.9 x 10-5 W.
For a solar-type star:
Sstar = 1.1 x 10-11 W.
For a solar-type star and a laser centered on the human visual response, the apparent magnitude may be expressed in terms of its intensity I:
m = -[19+2.5log(I)] (6)
Ilaser = 2.8 x 10-2
Substituting the above values for a
range of 10 light years, we find that:
mlaser ~ -15.
Photon Detection Rate
The signal photon detection rate is given by:
eta = photodetector quantum
Substituting the values in parentheses for a center wavelength of 550 nm, we find that the "signal" photon detection rate Nlaser:
Signal = 44,000 cpp (counts per pulse).
For a solar-type star, we find that the stellar background "noise" photon detection rate Nstar for an effective detection bandwidth of approximately 20 nm (human eye photopic response or optical detector response):
Noise = 6,000,000 cps (counts per second).
The "signal" is buried in the noise and the ratio between the "signal" and "noise" photons is approximately -20 dB. However, during each one nanosecond laser pulse, the SNR is positive and nearly 70 dB! This is the very important benefit obtained by searching for very short pulses in adjacent time slots corresponding to the expected pulse duration, even if the "signal" consists of only one or two detected photons per pulse. An additional critical benefit is that knowledge of "magic frequencies" is no longer required.
Figure 1 shows how the expected photon count changes with distance and receiver aperture size, based on the assumptions and analysis above. The photon count is shown for three ranges, defined by the spheres of 10, 100 and 1,000 light year radius. The bold type face denotes the photon count for "a great telescope", while the photon-counts in light face type will be more typical for amateur size receiving telescopes. Even at a range of 1,000 light years, sufficient photons per pulse can be received by a 10" (25.4 cm) telescope for the beacon to be considered detectable. It should be noted here that this diagram assumes that the overall photon detection efficiency through the atmosphere to the telescope receiver, including the quantum efficiency of the latter, is only a very conservative 1%. Thus, in practice, a 1018 W peak signal would be more than detectable by small telescopes. Especially, if fast detection and signal processing equipment exist for examining the statistics of the photon arrival times in each 1 ns time slot.
One of the major challenges of conducting Optical SETI by looking for fast beacon-type pulses, is the development of the ability to handle the vast data-rate and storage requirements. The following example illustrates the magnitude of the problem:
Star = G-Type
If all the stellar and sky background photons are counted and wideband optical filters are used, then the photon count rate can be very high, necessitating a large data storage capacity. In one hour, if every detected photon was recorded as a "1" in a spreadsheet, with each entry taking up 4 bytes, nearly 720 Mbytes of information would need to be stored. For a six-hour observation period, the amount of data to be stored would be about 4.3 Gbytes. For a very large telescope of 10-meter aperture, the data-rate would increase to about 78 Mbits/s, and the storage requirements would rise to 6.7 Tbytes of data per night! Thus, it can be argued that to undertake Optical SETI efficiently, we first need to develop low-cost, mass-data storage media and/or ultra-fast digital signal processing (DSP) that would remove the necessity to store data in its raw form. Fortunately, recordable CDROMs, otherwise known as CDROM-R, are rapidly becoming affordable, thereby allowing amateurs to store all the raw data obtained, should they so wish.
The issue of signal processing is one of the last issues to be addressed. To begin with, the sophistication of hardware and software used for this purpose is very dependent on the funding level available. For the moment, The Columbus Optical SETI Observatory intends to employ a very basic and somewhat insensitive scheme, which will be upgraded as funds become available. It is likely that LabVIEW for Windows will become an important tool for signal processing.
Figure 2 shows the main components of the observatory system, based around the Meade LX200 10" f/10 SCT. The telescope, and one computer, are housed in the 10 ft. PRO-DOME which is manufactured by Technical Innovations, Inc.. The dome is presently situated at the rear of the author's home, however, the intention is to move it to the roof of the garage when funds become available. Figure 3 shows the cable connections between the control room and the dome. The connections are as follows:
Because of the various control functions and data acquisition requirements, several PCs operating under Windows 95 will be employed. The multi-tasking capabilities of Windows 95, even on Pentium PCs, are insufficient to handle the demands of the observatory. So, dedicated computers will be used for the different functions. This has the added benefit of enabling each computer's video display to be converted to NTSC and stored on video cassettes. Figure 4 shows the computer control and signal processing system.
Figure 5 shows the video system employed to capture the video data. A quad video switching system allows many separate video signals to either be recorded together onto tape, or to be recorded onto tape in a sequential manner. A pulse stretcher is used to increase the detected pulse duration so that it can be reproduced on an audio system. The Photon "Snow" Converter allows a video display of photon arrivals to be converted into a standard NTSC video signal. Presently, it is planned to put an audio version of the photon count onto VHS cassettes via the right-hand channel of the hi-fi stereo tracks. It is also planned to store data in digital form, via the left-hand hi-fi audio track, rather than placing the digital data onto the video track, since this is occupied by quad video signals.
For observatory use only, the raw data may be archived at a much higher rate by exclusively reserving the video channel of a separate VCR system or by using a writable CD-ROM drive. The cost of writable CD-ROMs is declining quickly, it is now becoming much more economical to consider storing raw, unprocessed data on CD-ROMs. If a market developed for the raw data, the CD-ROMs could then be replicated, thereby enabling others to make use of their own signal processing algorithms. The author here stresses that the design of the observatory is an ongoing activity and is consequently subject to considerable further refinement.
Figure 6 shows a view of the 10 ft. dome looking north. The control room is to its right and the garage is situated to the left. Presently, the dome's height is limited to 7 ft.. This prevents the satellite dish, seen in the background, from striking the dome. Eventually, after the dome is moved to the roof of the garage, the height inside the dome will be increased to about 10 ft.
Figure 7 shows a close-up view of the LX200 SCT, with a piggybacked Meade 2045D 4" SCT. The latter is used with the SBIG ST-7 CCD camera in order to maintain accurate telescope tracking. Dew shields and heaters are employed to mitigate condensation problems. Strapped to the side of the 10" SCT are infrared transmitters and receivers. These monitor the position of the dome slit and produce error signals that rotate the dome via means of its two motors, and synchronously with the azimuth rotation of the telescope. Behind the telescope can be seen the dome's shutter and the shutter's motor, the latter having yet to be wired up.
Application for non-profit status will be submitted by The Columbus Optical SETI Observatory in 1996. Efforts will be made to obtain private funding for the extension of the signal processing and data storage capabilities of the observatory. When funds become available, the intention is to relocate the dome to the roof of the garage, since this will considerably enhance the view of the sky. The author wishes to cooperate with others in the development of international standards for data acquisition, and storage and signal processing. This will facilitate the international exchange of Optical SETI data and allow for the standardization of data formats. It is also intended that at some point in the next year or two, the observatory, when on-line, will be accessible through the Internet. Stored data may be downloadable at other times. See the paper by Jan Soldan and Milos Nemcek elsewhere in these proceedings regarding alternative approaches to the automation of OSETI observatories. Their approach will also provide connection with the Internet thereby giving remote access to data. But, additionally, it will permit the remote control of the Meade LX200 SCT.8.
* These addresses have changed: