≡ Menu

Pulsar Navigation: Mining Our Datasets

Science fiction dealt with interstellar navigation issues early on. In fact, Clément Vidal’s new paper, discussed in these pages yesterday, notes a George O. Smith story called “Troubled Star,” which originally ran in a 1953 issue of Startling Stories and later emerged as a novel (Avalon Books, 1957). Smith is best remembered for a series of stories collected under the title Venus Equilateral, but the otherwise forgettable Troubled Star taps into the idea of using an interstellar navigation network, one that might include our own Sun.

The story includes this bit of dialogue between human and the alien being Scyth Radnor, the latter explaining why his civilization would like to turn our Sun into a variable star:

“We use the three-day variable to denote the galactic travel lanes. Very effective. We use the longer variable types for other things – dangerous places like cloud-drifts, or a dead sun that might be as deadly to a spacecraft as a shoal is to a seagoing vessel. It’s all very logical.”

“…you’re going to make a variable star out of Sol, just for this?”

Well, why not, in Scyth Radnor’s view — after all, what’s one star in a galaxy-spanning navigation network? From our point of view, distant pulsars make for less local disruption, and as we saw yesterday, navigation by X-ray millisecond pulsars is already undergoing testing.

Image: Our early experiments, described yesterday, explore how we might use millisecond X-ray pulsars (MSPs) to provide autonomous spacecraft navigation. Credit: Astrowatch.net.

Visualizing a Pulsar Navigation Network

Using millisecond X-ray pulsars (MSPs) for galaxy-spanning navigation raises more than a few questions, especially when we try to predict what an artificial pulsar navigation system might look like to outside observers. If we are willing to posit for a moment a Kardashev II-level civilization moving between stars at relativistic velocities, then we would make as one of our predictions that such a system would be suitable for navigation at such speeds. In following the predictive model of Vidal’s paper, we would then check through our voluminous pulsar data to see how such a prediction fares. The answer, in other words, is in our datasets, and demands analyzing the viability of pulsar navigation at high fractions of c.

To my knowledge, no one has yet done this, making Vidal’s paper a spur to such research. The key here is to make predictions to see which can be falsified. But a quick recap for those just coming in on the discussion. What Vidal (Universiteit Brussel, Belgium) offers is an examination of millisecond X-ray pulsars as navigational aids, of the sort we’re already beginning to exploit through experiments via NASA, Chinese efforts and studies at the European Space Agency.

Specifically, the idea here is to develop a methodology for studying cases where astrophysical phenomena may have as one proposed explanation an extraterrestrial technology. Vidal also wonders whether we might find SETI implications even if a fully natural network like this were simply put to work by civilizations more advanced than our own, using it as we might wish to do.

Image: From Vidal’s paper, Fig. 4. Caption: A three-dimensional position fix can be obtained by observing at least three pulsars. Given three well-chosen pulsars, there is only one unique set of pulses that solve the location of the spacecraft (SC). Figure adapted from Sheikh (2005, 200).

Vidal is hoping to make predictions that are testable against our accumulating data to assess the likelihood of natural and artificial explanations, with pulsars as the case in point. The paper examines the kinds of predictions we would want to weigh against available data in a program the author calls SETI-XNAV. Whatever conclusions it reaches, such a program would improve our knowledge of pulsars themselves and our techniques at using them, possibly leading to our augmenting existing resources like the Deep Space Network with XNAV capabilities.

The author’s approach assumes that subjecting decades of data to analysis will teach us much about pulsar navigation as well as future SETI efforts:

Scientifically, SETI-XNAV is a concrete ETI hypothesis to test. The data is here, the timing and navigation functionalities are here. Historically, the suspicion of artificial canals on Mars triggered space missions to Mars and developed knowledge about Mars. Similarly, the project to try to decipher any potentially meaningful information in pulsar’s signals… could lead to the development of tools and methods that can be used for any future candidate signal.

That, of course, would augment the SETI effort as we expand into Dysonian SETI and the examination of possible engineering as the explanation for enigimatic astrophysical observations. If we assume a galaxy with a completely natural navigational system of this power, then we can imagine other civilizations putting it to use. Thus MSPs are likely to be standards in timekeeping and navigation for all putative civilizations in the Milky Way.

The Landscape of Prediction

Millisecond pulsars account for perhaps 10% of known pulsars, and as I mentioned yesterday, they appear to be distributed isotropically in the galaxy, a contrast to the rest of the pulsar population, which appears more concentrated in the galactic disk. MSPs offer numerous advantages from a navigational standpoint given that, according to Vidal, they are more than 100,000 times more stable than normal pulsars. Timing noise, an irregularity found in normal pulsars, and so-called ‘glitches’ (abrupt changes in rotation speed) are less frequent in MSPs. The latter are also associated with lower velocities than the other 90 percent of the pulsar population.

From the standpoint of artificiality, Vidal breaks the possibility terrain down seven ways (this is drawn from the paper’s Figure 1):

0 – Natural. All pulsars are natural. We are just lucky they provide stable clocks and an accurate navigation system

1 – Pulsars as standards. All pulsars are natural, but ETIs use them for timing, positioning and navigation purposes. Communication is galacto-tagged and time-stamped with a pulsar standard

2 – Natural and alterable. Some ETIs have the technology and capability to jam, spoof or interfere with a natural pulsar positioning system

3 – Artificial MSXP for navigation. Only a few millisecond X-ray pulsars have been modified by ETI for galactic navigation and timing purposes

4 – Artificial MSXP for navigation and communication. Only a few MSXPs have been modified by ETI, for navigation, timing and communication purposes

5 – Artificial pulsars. All pulsars are artificial. ETI build them, even the new ones, by intentionally triggering supernovas

6 – Artificial pulsars for us. All pulsars are artificial. ETI build them and they are currently sending us Earth-specific messages

The point here is telling for Dysonian SETI in general. We have established pulsar formation models that seem to work. To establish a program of XNAV-SETI, examining our storehouse of pulsar data, we do not need to challenge it.

But as we have learned more about pulsars over the years, we have learned that there is no unified pulsar model that explains the variety we have seen among this population. We can look toward understanding what MSPs are doing by asking what new hypotheses explain this rich set of observations.

The wide range of Vidal’s seven scenarios makes his case straightforward: “…we do not necessarily need to contradict existing pulsar models to entertain the possibility that ETI might be involved.” The issue then becomes, Vidal adds, to make and validate new predictions.

Emergent Questions

Yesterday I mentioned a recent paper examining radio pulsars in a SETI context. It was Chennamangalam, Siemion, Lorimer & Werthimer, “Jumping the energetics queue: Modulation of pulsar signals by extraterrestrial civilizations,” New Astronomy Volume 34, January 2015, pp. 245-249 (abstract). The paper examines the possibility of pulsars as ‘naturally occurring radio transmitters’ onto whose emissions information has been encoded. Vidal likewise thinks about millisecond X-ray pulsars in the context of possible information content, noting that Carl Sagan pondered studying pulsar amplitude and polarization nulls as far back as 1973.

It might be argued that communications signals would likely be compressed, making decoding extremely problematic, but Vidal’s point here is that navigational systems differ in fundamental ways from communications systems. Navigational signals should be more regular and easier to process than highly modulated signals with communications intent. If we are looking for content grafted onto the navigational signal, we can bring to bear the entire SETI toolkit, perhaps examining pulsar data in light of delay-tolerant networking and discontinuities in connectivity.

We move back into the area of predictions. World clocks on Earth are regularly re-synchronized, just as the time on global positioning satellites is synchronized through methods Vidal discusses, using a control segment that communicates with a satellite segment. Can we observe anything like this in our pulsar data? The author frames the matter this way:

The fastest and most stable MSPs might constitute such a control segment, to which the other pulsars would synchronize. Concretely, we could look for time correction signals broadcasts (that exist in GNSSs [Global Navigation Satellite Systems]), or synchronization waves. For example, synchronization might occur first on pulsars nearest the putative control segment and then diffuse to further away pulsars. This could be investigated via rare MSP glitches, or other remarkable features, such as giant pulses in MSPs.

Synchronization between MSPs would be evidence for a distributed solution on an interstellar level.

Other questions to explore: Do we find that MSPs further away from the galactic plane are more powerful than those closer in, potentially designed for low-density regions of the galaxy? Is MSP distribution random or does it show a pattern fitting the needs of galactic navigation? Estimates of the number of MSPs needed to navigate the entire galaxy might be contrasted with astrophysical predictions of the MSP population, currently estimated to be between 30,000 and 200,000. This one, of course, is tricky: We can only derive a theoretical lower boundary.

How MSPs form and evolve is fruitful ground for inquiry, given that some scientists have argued that the most commonly cited scenario for MSP evolution does not produce the X-ray MSP population we see. It is hard to see how an MSP in a non-binary system can maintain its spin without degrading over time, making the single MSP ground for study. Thus another round of prediction is possible. Single MSPs, those without an energy source, may simply be non-working parts of the network. Do we see redundancy between single and binary MSP coverage, given that binary MSPs are likely more reliable over long time periods?

What Vidal calls SETI-XNAV makes a significant departure from conventional SETI in the sense that it is not localized around a single star, but rather involves a search for a distributed signal that exists in the form of a navigation system, one either established by extraterrestrial engineering or simply relying on a natural phenomenon to pursue its own activities. That we can begin to use millisecond X-ray pulsars as navigation standards implies that more advanced civilizations have done so. Thus SETI-XNAV as constructed in this paper intends to survey the testable predictions against which we can run our expanding dataset on pulsars.

…all pulsars could be perfectly natural, but we can reasonably expect that civilizations in the galaxy will use them as standards… By studying and using XNAV, we are also getting ready to receive and send messages to ETI in a galactically meaningful way. From now on, we might be able to decipher the first level of timing and positioning metadata in any galactic communication.

But I would also emphasize that making testable predictions about pulsar navigation also exercises our skills at analyzing future astrophysical data that may prove enigmatic. That, in and of itself, is a useful contribution in this era of KIC 8462852 and ‘Oumuamua.

The paper is “Pulsar positioning system: a quest for evidence of extraterrestrial engineering,” published online in the International Journal of Astrobiology 23 November 2017 (abstract / preprint). See also Vidal, “Millisecond Pulsars as Standards: Timing, Positioning and Communication,” Proceedings IAU Symposium No. 337, edited by P. Weltevrede, B. B. P. Perera, L. Levin Preston, and S. Sanidas. Jodrell Bank Observatory, UK (2017). Preprint available.


Comments on this entry are closed.

  • Alex Tolley January 16, 2018, 16:04

    Given that K2 civs might be very old and have long since disappeared (e.g. cf Forbidden Planet and the Krell) it might also be worth extrapolating back through time the position of these MSP to see if their positions were less random at some point in the past. The Gaia data might be of help here.

  • J. Jason Wentworth January 16, 2018, 21:53

    I wonder if pulsars–whether natural, “altered natural,” or artificial objects–would necessarily be the *only* navigational reference points (just as our spacecraft often use both the nearby Sun and the distant star Canopus [due to the large Sun-spacecraft-Canopus angle] as navigational references), and:

    Duncan Lunan pointed out that for interstellar probes or starships, the galactic center would be a most useful reference point. While it can only be seen at radio (and perhaps infrared?) wavelengths from our area of the Milky Way due to the light-blocking clouds of interstellar dust, it might also be useful as an optical navigational reference for relativistic interstellar space vehicles. At such velocities, the emissions would be Doppler-shifted into the visible wavelengths, and it as well as pulsars could be used as navigational references.

    • Harold Shaw January 17, 2018, 14:06

      The relative position and orientation of an MSP to the galactic center could be very useful information. A lot of that information could be gathered at the MSP and, assuming a way to modify a MSP’s signal, transmitted as part of its identification signal.

  • J. Jason Wentworth January 16, 2018, 21:59

    ADDENDUM: I forgot to add (within the paired asterisks):

    “At such velocities, the emissions *(for spacecraft moving generally toward the galactic center)* would be Doppler-shifted into the visible wavelengths, and it as well as pulsars could be used as navigational references.”

  • Michael Fidler January 16, 2018, 23:19

    Some out of the box thinking ;-)

    The only way that this would work is if the system for communication would be using superluminal longitudinal waves. (Tesla Scalar waves)

    Particle Motion in Longitudinal Waves. 11*
    Superluminal and Luminal Waves
    E. T. Rowe

    “In this paper the treatment of the motion of a particle in a longitudinal wave is extended to the case of superluminal and luminal waves. Particle orbits are given in both closed and expanded forms. The results given here and in Part I are used extensively in the treatment of emission by particles in longitudinal waves which is to be given in a later paper. The treatment of the orbit presented here is important in that no approximation is made and so emission from particles
    in very strong plasma waves can be explored.
    The wave strength for a subluminal wave ro in the case of pulsars was briefly considered in Part I where it was found that it is between unity and 106 . The parameter roo, defined in Section 2, takes similar values. The development of a large amplitude coherent plasma wave in a pulsar magnetosphere, probably during the breakdown of the polar gap, needs to be considered in detail if this range of values is to be narrowed”.


    What are Scalar Waves?

    There is a wide confusion on what are scalar waves in serious and less serious literature on electrical engineering. In this paper we explain that this type of waves are longitudinal waves of potentials. It is shown that a longitudinal wave is a combination of a vector potential with a scalar potential. There is a full analogue to acoustic waves. Transmitters and receivers for longitudinal electromagnetic waves are discussed.


    The recent idea is that ions would work best to receive and transmit superluminal longitudinal waves but both the receiver and transmitter may be based on three ion (Plasma) interferometers, the transmitter being three separate ion magnetosphere (Plasma) pulsars.

  • Marshall Eubanks January 17, 2018, 12:44

    In many ways pulsar navigation is very analogous to GPS navigation. In both cases you have a set of remote clocks sending out timing signals – encoded radio (for GPS) and simple pulses (for pulsars). In both cases, you are not really doing absolute navigation, but are referencing your current position to a geodetic system set up previously. In the case of GPS, you cannot count on the spacecrafts having the same clock as you do, so you have to observe at least 4 satellites to get your position and the time offset between you and the GPS network (which synchronizes the satellite clocks to make navigation easier). The GPS network also encodes information about the positions of its satellites in the timing stream, so (for meter level positioning) you don’t to worry about modeling the satellite ephemeris.

    None of that information is available of course for pulsars, but, at least for the ones that are not in orbit about another star, their ephemeris is fairly easy to determine (a position and a velocity should suffice for periods << 1 million years), and millisecond pulsars are stable enough that most of the time you can get by with a quadratic clock model (pulsars all have a spin-down rate which must be observationally determined, and so pulsar clocks require a quadratic, not just a linear, error model). All of this modeling has to either be done before you leave on your voyage, or updates have to be sent to you en route (with substantial delays over interstellar distances).

    It seems like pulsar timing could be done with a good spacecraft clock and three pulsars, but in practice due to time dilation you won't know your clock's rate and offset unless you know your speed over the entire journey to the time of measurement, and so it would be better to assume the use of 4 pulsars (so you can determine your clock offset even if you have to reset your clock or lose track of your velocity as a function of time).

    But, even that's not sufficient, as pulsars are all subject to glitches (step function like changes in offset and rate, thought to do be due to "starquakes" in the pulsar crust). Glitches have to be determined observationally, which (if you are light years from home) means that you should be observing 5 or even 6 pulsars simultaneously. With 5 pulsars you could recover from one glitch and a clock reset, or two near-simultaneous glitches, and with 6 you could recover from two near-simultaneous glitches and a clock reset, which would make for a pretty robust navigation system.

    These considerations are not that important when you are a few light-hours away from home, as you could just wait for a new navigation update, but would become crucial with relativistic travel over multiple light years.

    • Mark Zambelli January 23, 2018, 6:42

      I don’t know enough about the technicals here so forgive me for asking, but wouldn’t it just be better to constantly monitor dozens of these MSP-beacons for error correcting redundancy? (For any ETI advanced enough to be utilizing this kind of positioning system this would seem to be well within their capabilities (processing, etc)), or would this not be advantageous?

  • Harold Shaw January 17, 2018, 14:50

    I think cases 5 and 6 would stand out. Case 5 predicts a busy galaxy. If case 5 is confirmed, then our discussions on METI must fundamentally change, become much more serious. Case 6, grab a towel.

    Cases where only a few MSPs are altered will be very challenging to find. For a galaxy spanning PPS, then the few altered likely correlate to the MSPs most fundamental to galaxy wide coverage. A model for PPS will be crucial. I would keep intergalactic possibilities in mind when building a PPS for interstellar possibilities. I think we could build a high fidelity model for a PPS, be confident that finding MSPs ideally situated is meaningful.

    For a local PPS, I think it will be more difficult to model. Where we know the shape of the galaxy, we would have to guess at the local PPS shape.

    To interrupt signal, we will need a better understanding of how MSPs generate their signal. A possible proportion to consider, signal dependence on internal mechanics or surface mechanics; internal mechanics are more expensive than surface mechanics to influence. If a MSP’s signal is overwhelmingly dependent on overall internal mechanics it may be prohibitively expensive to influence the natural signal. This would make placement of artificial MSPs more important. Another case that ideally placed MSPs would be meaningful.

    If a MSP’s signal depends enough on surface mechanics and those mechanics included generating a unique elemental signal, then bombarding a MSP with an accepted element could be used to build a signal. The infrastructure for this isn’t magical. That being said, would an orbiting object that interrupts the beam be more economical?

  • Harold Shaw January 17, 2018, 15:08

    Case 4.0 with a local almanac could be so productive, that evidence of case 5 would imply intergalactic demand. Case 6 should come with a warning. In the case of 6, I hope for a cat video. Humanity would let out the most anxious of laughs and be ok.

  • Mark Zambelli January 23, 2018, 6:50

    Ignoring the MSP-SETI implications and just considering the navigation side of things… how well do MSPs fare compared to using Quasars as your beacons?

  • ljk May 11, 2018, 11:15

    Hubble captures first image of surviving companion to a supernova

    And it seems that the companion star caused the explosion.

    By Jake Parks | Published: Thursday, April 26, 2018