We often conceive of SETI scenarios in which Earth scientists pick up a beacon-like signal from another star, obviously intended to arouse our attention and provide information. But numerous other possibilities exist. Might we, for example, pick up signs of another civilization’s activities, perhaps through intercepting electromagnetic traffic, or their equivalent of planetary radars? Even more interesting, as Brian McConnell speculates below, is the idea of listening in on a galactic network that contains information not just from one civilization but many. As Centauri Dreams readers know, McConnell and Alex Tolley have been developing the ‘spacecoach’ concept of interplanetary travel, discussed in the just published A Design for a Reusable Water-Based Spacecraft Known as the Spacecoach (Springer, 2015). It’s a shrewd and workable way to get us deep into the Solar System. Today McConnell turns his attention to a SETI network whose detection could offer a big payoff for a young civilization.

by Brian McConnell


With the revival of SETI funding, it’s interesting to contemplate what we might find if SETI succeeds. One possibility that is especially tantalizing is that first contact would not be with an individual civilization but rather a large scale network of civilizations that is organized not unlike the Internet. This is not a new idea (Timothy Ferris and others have explored this concept) but it is one that should be considered seriously. Assuming that communicative civilizations are commonplace, a big if of course, a decentralized or mesh network will be the most time and energy efficient way for them to organize their communications.

Consider the energy cost of sending a unit of information from one edge of the galaxy to the other (~ 100,000 light years) via direct means versus a peer-to-peer relay system. The savings ratio can be estimated as:


If communicative civilizations are separated by an average distance of, say, 1000 light years, the energy cost of sending a unit of information across the galaxy via relay will be about 1/100th that of direct communication. The energy requirement per link drops off by the ratio of (daverage / wgalaxy)2 but as more hops are required with shorter links, the overall energy requirement drops by daverage / wgalaxy. This is an approximation, but it highlights the order of magnitude improvements in economy, and suggests that if communicative civilizations are widespread, energy economics and other considerations will favor this type of arrangement.

Reliability and redundancy are another important feature of a mesh network. When sending information across such great distances, and with such long transit times, a sender may want to protect especially important information against loss or corruption by sending it repeatedly or by sending it via multiple paths between endpoints. This technique can virtually guarantee that information is eventually transmitted even if the network is damaged, even without the use of sophisticated forward error correction codes. This isn’t to say that an extraterrestrial intelligence will copy the TCP/IP protocol, but it’s safe to assume that someone who is sophisticated enough to build an interstellar communication link will probably be familiar with the characteristics and benefits of decentralized mesh networks.

There will also be benefits to receivers, especially newcomers, as contact with one node will be the same as contacting many nodes, since any node in the network can function as a relay for others. The cost of joining the network is also reduced, as a new node need only establish communication with its nearest neighbors, and can relay messages to and receive information from any other site on the network. Such a network would not merely be a communication system, but also a long term repository of knowledge as important information from long dead civilizations could continue to circulate throughout the network in perpetuity.


Image: The Milky Way as seen from the mountains of West Virginia. Could the galaxy be filled with the traffic of networked civilizations? Credit: ForestWander.

Fermi Implications

The existence of such a system might also help explain the Fermi Paradox, as the most energy efficient mode of operation, in terms of detecting new civilizations, will be for each node to concentrate its detection efforts on its immediate neighborhood using a listen and reply strategy. There would be little point in building powerful omnidirectional beacons that are detectable over great distances, as they would cost far more energy to operate, would have to wait millennia for a response, and would be plagued by duty cycle issues. Better for peripheral nodes to listen for microwave leakage from nearby civilizations as they develop early communication technology, and then target those for active communication soon after they are detected. This sort of strategy would be cheap both in terms of energy and the number of radio-telescopes required at each node in the network, and would offer a high probability of success in detecting new nodes just as they become active, while not wasting energy by transmitting in the blind.

An important point to consider here is that an emerging technological civilization would become detectable independently of any intent to attempt interstellar communication. Indeed here on Earth, the vast vast majority of energy expended on electromagnetic signaling has been for purposes other than Active SETI. It seems likely that most technological civilizations would go through a period where they are microwave bright, even if they later go dark due to transitioning to other technology, fear of ETI contact, etc.

As we would just now be detectable to nodes within about 100LY (80LY is probably a better estimate), we would just now expect to be receiving a response from a node within 40-50LY. It’s possible that rapid changes in atmospheric spectra, as Earth has experienced with the sudden increase in carbon dioxide, might also serve as a early tripwire for attempting active communication, but those could be ambiguous signals with natural explanations like volcanism, whereas a sudden spike in monochromatic microwave transmission points definitively to a technological origin. Viewed from the network’s perspective, this decentralized strategy would enable detection of new sites with the least energy expenditure and the shortest possible lag time between detection and active communication, with the added bonus feature that the first nodes to establish contact could relay stored information from nodes far beyond the initial radius of communication. On the other hand, if the average distance between nodes is large, it may be a long time before the nearest nodes are aware of us, and it may be hard for such a network to become established in the first place.

Choice of Encoding Schemes

Another interesting aspect of a long running galactic communication system is that there will be a natural selection of sorts that favors the message encoding schemes that are most likely to be mimicked. The selection pressure in this case will favor an encoding scheme that is broadly comprehensible (easy to understand the basic design pattern) and flexible (able to accommodate many different types of information via that framework). A transmission that is extremely difficult to parse, for example because of strong encryption or sophisticated forward error correction codes, is less likely to be mimicked than one whose basic design pattern is comprehensible to many receivers, even if it is less than optimal in terms of capacity or error resistance. This leads to the fittest message being more likely to replicate (be mimicked in retransmission) than its competitors. This is also an incentive for civilizations wishing to project influence through remote communication to design messages that peer sites will want to and be able to copy.

The point is not to speculate about what would be in such a message, but how it is organized at a low level. To build a mesh network that can handle many data types, you don’t need a very sophisticated message format, even if some of the data types sent within the message are extremely complex or difficult to comprehend. Typically you break a large amount of data, be it a file or communication stream, into smaller predictably organized subunits which are labeled with metadata, which might include:

  • a frame or packet number : identifies a message segment’s position within a collection, file, stream, etc
  • a collection or file number : to identify a larger grouping of frames, pages, packets, etc
  • an author or sender number : to identify the author or sender of a particular segment.
  • a receiver number : to identify the intended recipient, if any
  • a content identifier : to identify the type of content represented by the frame or packet
  • a blob of data, or payload, that is described by the above metadata

While one could design more complex schemes, the above defines the minimal set of metadata needed to describe something like a mesh network or file system with many files, authors and varying file types. What someone decides to convey with such a system is a different matter entirely, but a mesh network in its simplest form consists of a long chain of | meta data | blob of data | meta data | blob of data | segments with obvious repeating structures.

Implications for SETI

While the basic design pattern of a system like this can be rather simple, it will be capable of delivering data that varies widely in content type and “difficulty level”, and also offers a high degree of durability (important message fragments can be resent out of sequence or sent via multiple paths). Some content types such as rasterized or bitmapped images will probably be nearly universally understood due to their utility in astronomy and space photography, while others that are based on advanced math may be unrecognizable to many recipients. It’s not unlike DNA, whose basic encoding scheme has just four letters, yet can encode for something as simple as an isolated protein or as complex as a human being. That’s one of the interesting characteristics of the fittest message — it should be easy to parse at a low level, yet capable of conveying data types representing a wide range of complexity.

All of this suggests that SETI surveys should be concentrating a portion of their observing time on nearby targets. This also suggests that a large scale network will probably need to find us before we can find it, but will also be relatively easy to spot once it does. This doesn’t exclude other possibilities, and indeed SETI should be trying many strategies in parallel, from looking for distant beacons to Bracewell probes.

Should we encounter a network like this, the implications of that would be nothing short of staggering because of the volume and variety of information that could flow through a system like this. It’s possible that much of that communication will be over our heads. On the other hand, the quasi Darwinian selection pressure on message formats may favor those that are broadly comprehensible, or at least contain elements like rasterized photos that virtually any astronomically communicative receiver can understand, including us.