Journal of the British Interplanetary Society 36:490-495 (November, 1983)
Note: This web version is derived from an earlier draft of the paper and may possibly differ in some substantial aspects from the final published paper.
In this paper we shall demonstrate that, in most cases, interstellar spacecraft are actually superior to electromagnetic wave propagation for extrasolar exploration and communication, and as a focus for SETI research. We present arguments favouring interstellar communication via messenger probe, and critically review more than a dozen traditional objections.
In devising a figure of merit to permit ranking of proposed solutions to the problem of interstellar communication, based on the Principle of Economy, the relevant parameters to be examined are (1) information transmitted from sender to recipient and (2) distance between transmitter and receiver; the parameters to be minimised are (3) energy required for information transmission and reception and (4) time required for reception of the information. From these criteria the simplest interstellar communications figure of merit is defined as K (bits) (metre/sec)/joule, or dimensionally, the information density of momentum.
The Project Cyclops study  evaluated specific proposed optical, infrared and microwave systems for interstellar communication and found the microwave system superior to either of the others. The figures of merit for competing state-of-the-art designs labelled Optical A (Ko), Infrared A (Ki), and Microwave A (Km) are compared quantitatively by calculating the merit ratios (using data from Table 5-3 of the Cyclops report: DKm/o = Km/Ko 9000 and DKm/i = Km/Ki = 6000. Thus the information density of momentum in microwave signals can be 9000 times greater than for optical signals and 6000 times greater than for infrared signals, a tremendous relative advantage.
where H is Shannon's information function. This is a maximum for p1 = p2 = …, = pn = 1/n, in which case H = log2(n). As a theoretical limit, one or more sets of quantum energy levels may be used as information markers in solid-state artifacts . These sets of levels must he within the interval (0, E) where E = Mc2, the mass-energy of the information storage system. If energy levels can at best be measured only to an accuracy DE, then at most N = E/DE nonzero levels may be distinguished within a single set spanning the interval (0, E). Including the zero value, n = 1 + N energy levels may be distinguished, so in this case H = log2(n) = log2(1+N). If no more than n = 1 + N/j energy levels are occupied in each of j sets, with each set spanning an interval (0, E/j), then at most H = j log2(n) = j log2(1+ N/j) bits may be represented by the entire storage system. For j = N with markers spanning the interval (0, DE), maximum storage is achieved and Hmax = N log2(1+ N/N) = N bits. The Uncertainty Principle defines the minimum energy measurement accuracy DE = h/Dt, where h is Planck's constant and At is the uncertainty in the duration of the measurement, so the maximum storage capacity of the messenger probe Np = Hmax = N = E/DE = Mc2Dt/h bits. The message travels at a velocity Vp < c, assumed invariant. The energy required to accelerate the information package up to Vp during transmission, plus an equal amount for full deceleration to achieve reception, plus rest energy, is Ep = 2 Mc2 (1-Vp2/c2)-1/2. Thus the figure of merit for messenger probes Kp = NpVp/Ep = 1/2 (Dt/h)Vp(1-Vp2 /c2)1/2 .
In the case of Vp = Vopt, the information density of momentum for radio waves exceeds that of probes by a slight factor DKe/p = 4 for S/N = 1, and DKe/p = 40 for S/N = 1000. For S/ N < 0.2, interstellar probes are actually superior because DKe/p < 1. Recall that all these ratios are far less than the DKm/o = 9000 for microwave over optical systems and the DKm/i = 6000 for microwave over infrared systems computed earlier - and that optical and infrared are still considered plausible frequency bands for SETI work by some.
Thus for interstellar probes and radio waves that might be employed for SETI, DKe/p is approximately of order unity. That is, electromagnetic and material artifacts in theory can achieve a comparable information density of momentum.
From the above, it seems there exists no strong, purely physical basis for choosing between electromagnetic signals and material artifacts for interstellar missions of exploration and communication. Other logical selection criteria must be sought. The following arguments are advanced to show that in most cases probes are the modality of choice.
A negative cannot be proven, so a beacon operated for, say, a million years which receives no response permits few definitive conclusions regarding the galactic distribution of life and intelligence. Only probes can reliably and conclusively return information from every star system surveyed, sufficient to establish the truth or falsity of statistical assertions concerning ETI.
With probes, senders can install a permanent, self-repairing galactic SETI network at the cost of one self-reproducing interstellar probe [21, 23]. Probes can park in orbit for hundreds or even millions of years, awaiting discovery by emerging communicative cultures on suitable planets in the star system; alternatively, they may be programmed to execute multistar flyby trajectories until a communicative species is located, then enter into a dialogue with them at no further cost to the original senders.
With beacons, huge transmitters must be constantly supervised, serviced, and fed energy that the sending society could probably better use elsewhere. Beacons may radiate energy and information for millennia or longer without gaining any response or new information in return. This energy, since it was detected by no receiver, in essence was wasted and constitutes pure economic loss for the transmitting society.
A major advantage of probes during acquisition is their ability to place a high-strength signal in the-immediate vicinity of the target planet. Such "local beacons" will immediately attract attention as unmistakable evidence of the existence of ETI. The extraterrestrial origin of the probe could be quickly and accurately verified to the satisfaction of all skeptics. Much like the current "evidence" for UFOs, a beacon signal could all too easily be suspect as a hoax, military diversion, or be explained away as only a very unusual natural phenomenon.
Contact via probes provides a potentially richer, deeper interaction than via radio waves. A probe's onboard memory elements may contain an appreciable fraction of the knowledge and culture of the sending civilisation. By comparison, electromagnetic message scenarios appear little better than sterile data swaps or massive data dumps with extraordinarily poor bit rates.
An interesting alternative suggestion is that the a radio message may consist of instructions to build a probe (or computer) which can then achieve the desired swift interaction [241. Unfortunately, there are numerous difficulties with this approach:
By comparison, all a remote beacon can do is keep signalling blindly. Such beacons offer no possibility of modifying the signal until at least one light-speed round trip travel time has elapsed. Unlike probes, radio waves cannot modify themselves in response to unexpected target characteristics.
On the other hand, if local technological activity is detected by an intelligent messenger probe, the device may initiate surveillance or contact without ever having to disclose the identity or whereabouts of its creators. The probe is also free to withhold any sensitive information concerning weakness or strengths of the senders. (The probe could even supply military misinformation, making the senders appear more formidable than they really are and so discouraging aggression). If the probe must make periodic reports this may be accomplished in a. manner virtually impossible to trace or to decipher, using relays through uninhabited star systems, multiple transmissions or dummy broadcasts, and prime-number or "trapdoor"  encryptation. Probes permit senders to retain control of the interaction; beacons donate that control to (unknown, possibly unfriendly) recipients.
Self-repairing probe data banks may preserve for geological timescales the knowledge and cultural heritage of the transmitting civilisation, a record of successful and failed survival strategies of incalculable value to an expanding technological species. In the event the sending society is destroyed or perishes for any reason, a self-replicating system (SRS) [27-28] could resurrect its creators, their natural ecology and civilisation, even the original home world from stored data using SRS terraforming methods .
Despite the numerous obvious advantages of probes for exploration and communication, over the years many objections have been raised on physical, economic, and logical grounds. Close examination reveals each objection is either incorrect, unfounded, or, in some cases, actually a good argument in favour of probes.
For example, to operate a gigawatt beacon for a million years costs about the same energy as sending out a fleet of 100 kg 10% c probes to each of the million target stars at the rate of one per year. During this time the beacon will probably elicit no response to its call and thus yields zero knowledge, whereas the probe fleet is certain to have returned comprehensive data on a million nearby solar systems even if no intelligent life is found. This seems more cost-effective.
Further, if the probe is self-replicating only one initial device need be sent out. For a payload mass of 100 tons  a single 10% c self-reproducing interstellar probe  can initiate a wave of galactic exploration for an energy investment of only 1020 J. This requires an expenditure of about 30 gigawatts over a period of 100 years (approximate travel time to the nearest stars), which could be entirely supplied by one 15 km-wide 10% efficient solar power satellite in Earth orbit, upon arrival, the self-replicating probe could be programmed to construct a gigawatt SETI radio beacon "for free" in the target solar system at a (militarily) safe distance from the home world, or to terraform local planets  in the target system by remote command as a prelude to colonisation by the senders.
At v = Vopt, Rm = 5.8 for a photon rocket; with a fusion rocket, Rm = 1.3 = 106. For a 100 kg interstellar probe this latter value gives a total starting booster mass of 130,000 tons, only about 50% larger than the carrier USS Nimitz and less than a third the deadweight tonnage of the largest seagoing tankers. The argument that - is enormous for v = 0.99 c  merely shows that it is impractical - and fortunately unnecessary - to travel this fast.
In any case, probes may be externally powered so mass-ratio is not a threshold parameter for interstellar flight. Further, the discussion in Section 2.3 shows that the minimum energies required to transmit information across interstellar distances via radio or material artifacts are of comparable magnitude.
Yet almost certainly any race capable of transmitting either radio signals or interstellar probes for SETI purposes must be far in advance of ourselves, possibly on the level of a Kardashev Type II civilisation (utilising a major fraction of the energy output of their sun). To launch a 100,000-ton vehicle on a one-way trip at 1 Earth gravity acceleration to a destination 100 light-years away requires an equivalent relative energy expenditure for a Type II civilisation as the launching of a few Saturn V rockets represented to American society a decade ago. Active interstellar exploration by advanced cultures will require commitment but hardly an outrageous sacrifice.
The energies required to send probes or to maintain a beacon are about the same, as are the requirements to receive information across interstellar distances via artifacts or radio waves, so this objection applies equally strongly to radio-SETI efforts. Buying the "good life"  is subjective and a cultural relative, and for many includes active cosmic exploration. The notion that the senders' home planets could be environmentally restructured for the same (energy) cost of launching a probe fleet ignores current estimates of 1024-1028 J to terraform Mars and Venus to Earthlike conditions [29, 32], as compared to 1020 J for the million-probe (self-replicating) fleet and almost 1023 J for a beacon.
The enormous advantage of a probe's ability to establish a rapid information exchange with local communicants immediately upon arrival (or any time afterwards) more than offsets any slight disadvantage of slower initial transmission speed. Probes eliminate the time wasted in waiting for radio messages to crawl back and forth between the stars.
If the average waiting time for communicative civilisations to emerge is of the same order as interstellar flight times (100-10,000 years), then the small temporary speed advantage of radio waves evaporates. On the other hand, in the unlikely event that the radio beacon signal is detected immediately upon arrival, there is still no time even for a single complete round-trip message exchange before a v = Vopt probe launched simultaneously with the beacon signal arrives.
It is equally easy to receive probes or photons, reception being the only relevant parameter from the point of view of the recipient. It has already been established that from the viewpoint of the sender, transmissions of beacon signals or probes are energetically equivalent exercises.
On Earth the first gas reaction jet and first rockets were built 2200 years and 600 years ago, respectively, pre-dating the invention of radio by many centuries. The Project Daedalus starship study  suggests that radio and interstellar flight technologies probably should be regarded as virtually simultaneous technical developments in interstellar communication and evolutionary time frames. And the transmission technology a potential recipient species happens to possess at a give moment in its historical development is arbitrary and thus largely irrelevant to the choice space of message senders.
First, all probes needn't be monitored simultaneously because usually they'll have nothing new to report. Bracewell  claims that one 100-metre dish should be enough to maintain a serial monitoring schedule with no data loss for the fleet. Spherical lens designs  also may permit many targets to be monitored simultaneously.
Second, if everyone is listening who is sending? If the antenna-builders send out probes they are absolutely certain of receiving some useful information. If they send no probes but merely listen with their Cyclopean array, they most likely will never hear anything and thus will accomplish nothing.
Third, the argument confuses senders with recipients. For the recipients (e.g. humanity) the major SETI options are (1) build radio telescopes to search for beacons or (2) build optical telescopes to search for artifacts. These alternatives appear equal in cost for an effective search. For the sending society (e.g., advanced ETI), the major options are (1) build and maintain a radio beacon transmitter or (2) build and launch a fleet of exploratory probes. These choices also appear roughly equal in cost for a credible acquisition effort.
Interstellar exploration is necessarily a long-term effect, so any probe detected undoubtedly is the product of a very mature technology, long developed and near-perfectly adapted for the task assigned to it. Also, obsolescence after launch  is irrelevant from the viewpoint of the recipient, who seeks only to detect the presence of an artifact in his solar system, and is unimportant from the standpoint of the sender, as long as the device performs its assigned tasks adequately though perhaps not optimally. If the probe is self-repairing, as seems likely, it can be continuously updated. reprogrammed or physically restructured to reflect any necessary improvements.
The Project Daedalus starship design study assumes a system reliability of 99.99% (the same as NASA's Project Apollo) over a nominal 50-year mission lifetime  This implies a survival probability of 99.8% after 1000 years and 81.9% after 100,000 years. To achieve this with foreseeable technology a "repair-by-repair strategy" is required in which failed components are repaired and then returned to service.
The notion that artifact searches necessarily imply action by the other party but not by us  arose in response to an early suggestion that recipients should wait for extraterrestrial probes to initiate contact [381. This strategy is not being proposed here, Rather, advanced ETI will send interstellar probes to promising star systems, but it may be necessary for potential recipients to actively search for these artifacts in order to initiate communication .
Extraterrestrial artifacts are most likely to be found in four major orbital regions, analogous to the radio "waterhole" frequencies . Each region subsumes many possible stable orbits; however, the stopping criterion for probe searches is conceptually similar to that defined for microwave SETI searches over a frequency/power/time/polarisation phase space. Given a search volume and minimum artifact size/albedo , specific instruments may be used to search for artifacts to any desired probability of completeness.
Quite the contrary, the search space for extraterrestrial artifacts is sufficiently broad to include many collateral targets of great astronomical interest. For example, a SETI programme for artifacts would overlap to some degree with natural satellite and asteroid searches, assist in the cataloguing of resources for future space manufacturing ventures, and would hasten the development of large, automated optical space telescopes useful in space debris tracking, planetary remote imaging, solar astronomy, stellar census surveys, faint object detection and classification, and cosmology.
Until recently [39-40, 32-43], few seriously considered the appearance of such artifacts or specific strategies to search for them. UFOs [44-46] are not interstellar probes, regardless of their observational status. We have little evidence to confirm or dispute the presence of self-replicating alien machines in the Asteroid Belt ; self-repairing or reproducing probes may not rare whether or not we find them, may be occupied solely with noninvasive activities, or may, like NASA scientists, be interested in SETI (search) but not CETI (communication).
In the search for extraterrestrial artifacts (SETA)  a special case of the Fermi Paradox is apparent: Why are They silent? One possibility is that the probe is currently in a state of dormancy pursuant to a standard sleep/wake cycle. Least probable is that it has become a derelict. It must be presumed that a still-functioning advanced extraterrestrial spacecraft could make its presence known to us at any time if it desired to do so. One explanation for its silence if it exists is that the artifact is still studying us - a 0.1 -metre or larger antenna located closer to Earth then Sun-Earth L1 could detect both carrier and programming for VHF and UHF television broadcasts .
Perhaps the more relevant question is: Why should They not be silent? It is anthropocentric to assume that an alien spacecraft entering the Solar System on a mission of reconnaissance or self-replication will feel obligated to announce its presence to us or to request permission to proceed. Most likely, the probe will simply ignore us and go on about its business. If we want to find it, we'll have to go looking.
SETI is not a zero-sum game. Artifact and radio beacon searches can be conducted entirely independently, although it is interesting that searches for radio beacons can provide indirect limits on the existence of probes in the Solar System. For example, the all-sky survey outlined by Wolfe et al  would provide an excellent independent observational limit on the minimum size of an Earth-Moon orbiting artifact maintaining its own local acquisition beacon. Specifically, a solar-powered artifact of size L ran operate a beacon with maximum continuous power aL2 LS and maximum gain 4pL2l2, where LS is the local solar constant, L is wavelength and a < 1 is the fraction of probe power diverted to the beacon. The minimum detectable power in the proposed Sky Survey mode is d = 3 x 10-23 W/m2 So the minimum detectable beacon probe is of size L = (dl2R2/ aLS)1/4 = 3.4 mm for l = 0.2 metre, R = 3.84 x 108 metres (Lunar orbit), and a =1. For a = 10-10, L = 1 metre, so the limit is insensitive to beacon power. Regular radio-SETI searches of various nearby stars might detect incoming messages directed at resident probes. Targeted radio listening searches of likely probe residence orbits  or the Asteroid Belt  could also be conducted in an eavesdropping mode to detect accidental electromagnetic leakage radiation.
If extraterrestrial life and intelligence exist, and if these ETI have ever engaged in, or presently are engaging in, interstellar exploration or communication, this most likely will involve the transmission of material artifacts. Some evidence of this activity may be apparent from within the confines of the Solar System and thus could be detected by a suitable observational effort [20, 39, 40].