Autonomous Spacecraft Navigation With Pulsars

Werner Becker1,2∗, Mike G. Bernhardt1 , Axel Jessner2

arXiv:1305.4842v1 [astro-ph.HE] 21 May 2013

1 Max-Planck-Institut für extraterrestrische Physik, Gießenbachstraße, 85748 Garching, Germany

2 Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany

March 13, 2013

Abstract surements taken by ground stations provide very

accurate information on the distance and the ra-
An external reference system suitable for deep dial velocity of the spacecraft with typical ran-
space navigation can be defined by fast spinning dom errors of about 1 m and 0.1 mm/s, respectively
and strongly magnetized neutron stars, called pul- (Maddè et al., 2006). The components of position
sars. Their beamed periodic signals have timing and velocity perpendicular to the Earth-spacecraft
stabilities comparable to atomic clocks and provide line, however, are subject to much larger errors due
characteristic temporal signatures that can be used to the limited angular resolution of the radio an-
as natural navigation beacons, quite similar to the tennas. Interferometric methods can improve the
use of GPS satellites for navigation on Earth. By angular resolution to about 25 nrad, correspond-
comparing pulse arrival times measured on-board ing to an uncertainty in the spacecraft position of
a spacecraft with predicted pulse arrivals at a refer- about 4 km per astronomical unit (AU) of distance
ence location, the spacecraft position can be deter- between Earth and spacecraft (James et al., 2009).
mined autonomously and with high accuracy ev- With increasing distance from Earth, the position
erywhere in the solar system and beyond. The error increases as well, e.g., reaching a level of un-
unique properties of pulsars make clear already to- certainty of the order of ±200 km at the orbit of
day that such a navigation system will have its ap- Pluto and ±500 km at the distance of Voyager 1.
plication in future astronautics. In this paper we Nevertheless, this technique has been used suc-
describe the basic principle of spacecraft naviga- cessfully to send space probes to all planets in the
tion using pulsars and report on the current devel- solar system and to study asteroids and comets at
opment status of this novel technology. close range. However, it might be necessary for fu-
ture missions to overcome the disadvantages of this
method, namely the dependency on ground-based
1 Introduction control and maintenance, the increasing position
and velocity uncertainty with increasing distance
Today, the standard method of navigation for in-
from Earth as well as the large propagation delay
terplanetary spacecraft is a combined use of radio
and weakening of the signals at large distances. It
data, obtained by tracking stations on Earth, and
is therefore desirable to automate the procedures
optical data from an on-board camera during en-
of orbit determination and orbit control in order to
counters with solar system bodies. Radio mea-
support autonomous space missions.
∗ [email protected]

1
 Possible implementations of autonomous navi- of position determination by comparing pulse ar-
gation systems were already discussed in the early rival times at the spacecraft with those at a refer-
days of space flight (Battin, 1964). In principle, ence location. Within the limitations of technol-
the orbit of a spacecraft can be determined by mea- ogy and pulsar data available at that time (a set
suring angles between solar system bodies and as- of only 27 radio pulsars were considered), Downs
tronomical objects; e.g., the angles between the showed that spacecraft position errors on the or-
Sun and two distant stars and a third angle be- der of 1500 km could be obtained after 24 hours
tween the Sun and a planet. However, because of signal integration. A possible improvement in
of the limited angular resolution of on-board star precision by a factor of 10 was estimated if better
trackers and sun sensors, this method yields space- (high-gain) radio antennas were available for the
craft positions with uncertainties that accumulate observations.
typically to several thousand kilometers. Alterna- Chester & Butman (1981) adopted this idea and
tively, the navigation fix can be established by ob- proposed to use X-ray pulsars, of which about
serving multiple solar system bodies: It is possi- one dozen were known at the time, instead of ra-
ble to triangulate the spacecraft position from im- dio pulsars. They estimated that 24 hours of data
ages of asteroids taken against a background field collection from a small on-board X-ray detector
of distant stars. This method was realized and with 0.1 m2 collecting area would yield a three-
flight-tested on NASA’s Deep-Space-1 mission be- dimensional position accurate to about 150 km.
tween October 1998 and December 2001. The Au- Their analysis, though, was not based on simula-
tonomous Optical Navigation (AutoNav) system tions or actual pulsar timing analyses; neither did
on-board Deep Space 1 provided the spacecraft it take into account the technological requirements
orbit with 1σ errors of ±250 km and ±0.2 m/s, or weight and power constraints for implementing
respectively (Riedel et al., 2000). Although Au- such a navigation system.
toNav was operating within its validation require- These early studies on pulsar-based navigation
ments, the resulting errors were relatively large estimated relatively large position and velocity er-
compared to ground-based navigation. rors so that this method was not considered to be
In the 1980s, scientists at NRL (United States an applicable alternative to the standard naviga-
Naval Research Laboratory) proposed to fly a tion schemes. However, pulsar astronomy has im-
demonstration experiment called the Unconven- proved considerably over the last 30 years since
tional Stellar Aspect (USA) experiment (Wood, these early proposals. Meanwhile, pulsars have
1993). Launched in 1999 on the Advanced Re- been detected across the electromagnetic spectrum
search and Global Observation Satellite (ARGOS), and their emission properties have been studied in
this experiment demonstrated a method of position great detail (cf. Becker, 2009a, for a collection of
determination based on stellar occultation by the comprehensive reviews on pulsar research). Along
Earth’s limb as measured in X-rays. This tech- with the recent advances in detector and telescope
nique, though, is limited to satellites in low Earth technology this motivates a general reconsidera-
orbit. tion of the feasibility and performance of pulsar-
An alternative and very appealing approach to based navigation systems. The present paper re-
autonomous spacecraft navigation is based on pul- ports on our latest results and ongoing projects in
sar timing. The idea of using these celestial this field of research. Its structure is as follows:
sources as a natural aid to navigation goes back After summarizing the most relevant facts on pul-
to the 1970s when Downs (1974) investigated the sars and discussing which pulsars are best suited
idea of using pulsating radio sources for interplan- for navigation purposes in § 2, we briefly describe
etary navigation. Downs (1974) analyzed a method the principles of pulsar-based navigation in § 3.

2
Pulsars emit broadband electromagnetic radiation stars are observable as pulsars if their spin axis
which allows an optimization for the best suited and magnetic field axis are not aligned. Having
waveband according to the highest number of bits co-rotating magnetic fields of B⊥ ≈ 109 –1013 G
per telescope collecting area, power consumption, and spin periods down to milliseconds they ra-
navigator weight and compactness. A possible an- diate broadband electromagnetic radiation along
tenna type and size of a navigator which detects narrow emission cones. If this radiation cone
pulsar signals at 21 cm is described in § 4. In § 5 crosses the observer’s line of sight a pulse of in-
we discuss the possibility of using X-ray signals tensity is recorded in the observing device (cf. Fig-
from pulsars for navigation. The recent develop- ure 1). The name Pulsar refers to this prop-
ments of low-mass X-ray mirrors and active-pixel erty. They have been discovered by their radio
detectors, briefly summarized in § 6, makes it very signals (Hewish et al., 1968). In source catalogs
appealing to use this energy band for pulsar-based their common abbreviation is therefore PSR which
navigation. stands for Pulsating Source of Radio, although
they have also been detected in other bands of the
electromagnetic spectrum meanwhile. Three dif-
2 The Various Types of Pulsars ferent classes of pulsars can be distinguished ac-
and their Relevance for Naviga-
tion
Stars are stable as long as the outward-directed
thermal pressure, caused by nuclear fusion pro-
cesses in the central region of the star, and the
inward-directed gravitational pressure are in equi-
librium. The outcome of stellar evolution, though,
depends solely on the mass of the progenitor star.
A star like our sun develops into a white dwarf.
Stars above ≈ 8 M⊙ undergo a gravitational col-
lapse once their nuclear fuel is depleted. Very
massive stars of more than about 30 M⊙ end up
as black holes and stars in the intermediate mass
range of about 8 to 30 M⊙ form neutron stars.
It is assumed that a neutron star is the result of
a supernova explosion, during which the bulk of
its progenitor star is expelled into the interstellar
medium. The remaining stellar core collapses un-
der its own weight to become a very compact ob-
ject, primarily composed of neutrons – a neutron Figure 1: Artist’s impression of a rotation-powered
star. With a mass of typically 1.4 M⊙ , compressed pulsar. The neutron star appears as a pulsating source of
radiation if the rotating emission beam crosses the ob-
into a sphere of only 10 km in radius, they are quasi
server’s line of sight. Averaging these periodic pulses
gigantic atomic nuclei in the universe. Because of of intensity over many rotation cycles results in a stable
their unique properties they are studied intensively pulse profile. Because of the timing stability of most
by physicists of various disciplines since their dis- pulsars, the arrival time of pulses can be predicted with
covery in 1967 (Hewish et al., 1968). very high precision, which is an essential requirement
Fast spinning and strongly magnetized neutron for a navigation system based on pulsar observations.

3
cording to the energy source of their electromag- residual fall-back disk (Trümper et al., 2010,
netic radiation. As we will see, only one class is 2013). However, their long-term timing be-
suitable for spacecraft navigation: havior is virtually unknown, which invali-
dates these sources for the use in a pulsar-
• Accretion-powered pulsars are close binary
based spacecraft navigation system.
systems in which a neutron star is accreting
matter from a companion star, thereby gain- Concerning their application for navigation, the
ing energy and angular momentum. There only pulsar class that really qualifies is that of
are no radio waves emitted from the accre- rotation-powered ones.
tion process, but these systems are bright in
X-rays. The observed X-ray pulses are due • Rotation-powered pulsars radiate broad-
to the changing viewing angle of a million band electromagnetic radiation (from radio
degree hot spot on the surface of the neu- to optical, X- and gamma-rays) at the ex-
tron star. These hot spots are heated by pense of their rotational energy, i.e., the pul-
in-spiraling matter from an accretion disk. sar spins down as rotational energy is radiated
The accretion disk and the accretion col- away by its co-rotating magnetic field. The
umn itself can also be sources of X-rays. amount of energy that is stored in the rota-
The spin behavior of accretion-powered pul- tion of the star can be estimated as follows: A
sars can be very complicated and complex. neutron star with a radius of R = 10 km and
They often show an unpredictable evolution a mass of M = 1.4 M⊙ has a moment of in-
of rotation period, with erratic changes be- ertia I ≈ (2/5) M R2 ≈ 1045 g cm2 . The rota-
tween spin-up and spin-down as well as X- tional energy of such a star is Erot = 2π 2 I P−2 .
ray burst activities (Ghosh, 2007). Although Taking the pulsar in the Crab nebula with
accretion-powered pulsars are usually bright P ≈ 33 ms as an example, its rotational En-
X-ray sources, and thus would give only mild ergy is Erot ≈ 2 × 1049 erg, which is compara-
constraints on the sensitivity requirements of ble with the energy released by thermonuclear
a pulsar-based navigation system, their un- burning of our sun in hundred million years.
steady and non-coherent timing behavior dis- The spin period of a rotation-powered pulsar
qualifies them as reference sources for navi- increases with time due to a braking torque
gation. exerted on the pulsar by its magneto-dipole
• Magnetars are isolated neutron stars with ex- radiation. For the Crab pulsar, the observed
ceptionally high magnetic dipole fields of up period derivative is Ṗ = 4.2 × 10−13 s s−1 ,
to 1015 G. All magnetars are found to have ro- which implies a decrease in rotational energy
tation periods in the range of about 5 to 10 of Ėrot = −4π 2 I Ṗ P−3 ≈ 4.5 × 1038 erg s−1 . It
seconds. RXTE and other X-ray observato- has been found, though, that the spin-down
ries have detected super-strong X-ray bursts energy is not distributed homogeneously over
with underlying pulsed emission from these the electromagnetic spectrum. In fact, only
objects. According to the magnetar model a fraction of about (10−7 − 10−5 ) Ėrot is ob-
of Duncan & Thompson (1992), their steady served in the radio band whereas it is roughly
X-ray emission is powered by the decay of (10−4 − 10−3 ) Ėrot in the X-ray band and
the ultra-strong magnetic field. This model (10−2 − 10−1 ) Ėrot in the gamma-ray band
also explains the X-ray burst activity observed (Becker, 2009b).
from these objects, but there are also alterna- There are two types of rotation-powered pul-
tive theories, which relate these bursts to a sars: (1) Field pulsars have periods between

4
 tens of milliseconds to several seconds and they are connected by an evolutionary pro-
constitute more than 90% of the total pulsar cess: It is assumed that millisecond pulsars
population. (2) About 10% of the known pul- are born as normal pulsars in a close binary
sars are so-called millisecond pulsars, which system, but their rotation accelerates as they
are defined to have periods below 20 mil- pass through a phase of accretion in which
liseconds. They are much older than nor- mass and angular momentum are transferred
mal pulsars, posses weaker magnetic fields from the evolving companion star to the
and, therefore, relatively low spin-down rates. pulsar (e.g., Bhattacharya & van den Heuvel,
Accordingly, they exhibit very high timing 1991). However, the fact that millisecond
stabilities, which are comparable to atomic pulsars are often found in binary systems
clocks (Taylor, 1991; Matsakis et al., 1997). does not affect their suitability for space-
This property of millisecond pulsars is of craft navigation as the binary motion can
major importance for their use in a pulsar- easily be accounted for in pulsar timing
based navigation system. Figure 2 clearly (Blandford & Teukolsky, 1976). Millisecond
shows that these two types of pulsars be- pulsars – also referred to as recycled pul-
long to distinct populations. Most likely, sars – were discovered by Backer et al. (1982)
and studied extensively in the radio band
by, e.g., Kramer et al. (1998). Pulsed X-ray
emission from millisecond pulsars was dis-
covered by Becker & Trümper (1993) using
ROSAT. However, only XMM-Newton and
Chandra had the sensitivity to study their X-
ray emission properties in the 0.5 − 10 keV
band in greater detail. The quality of data
from millisecond pulsars available in the X-
ray data archives, though, is still very inho-
mogeneous. While from several of them high
quality spectral, temporal and spatial infor-
mation is available, many others, especially
those located in globular clusters, are detected
with just a handful of events, not allowing,
e.g., to constrain their timing and spectral
properties in greater detail. From those mil-
lisecond pulsars detected with a high signal-
to-noise ratio strong evidence is found for
a dichotomy of their X-ray emission prop-
erties. Millisecond pulsars having a spin-
down energy of Ė ≥ 1035 erg/s (e.g., PSR
Figure 2: The P-Ṗ diagram; distribution of rotation-
J0218+4232, B1821−24 and B1937+21)
powered pulsars according to their spin parameters.
X-ray detected pulsars are indicated by colored sym- show X-ray emission dominated by non-
bols. The straight lines correspond to constant ages thermal radiation processes. Their pulse pro-
τ = P/(2Ṗ) and magnetic field strengths B⊥ = 3.2 × files show narrow peaks and pulsed fractions
1019 (PṖ)1/2 as deduced within the framework of the close to 100% (cf. Figure 3). Common for
magnetic braking model. these pulsars is that they show relatively hard

5
 glitches in their spin-down behavior, i.e., abrupt
increases of rotation frequency, often followed by
an exponential relaxation toward the pre-glitch
frequency (Espinoza et al., 2011; Yu et al., 2013).
This is often observed in young pulsars but very
rarely in old and millisecond pulsars. Neverthe-
less, the glitch behavior of pulsars should be taken
into account by a pulsar-based navigation system.
Today, about 2200 rotation-powered pulsars are
known (Manchester et al., 2005). About 150 have
been detected in the X-ray band (Becker, 2009b),
and approximately 1/3 of them are millisecond pul-
sars. In the past 30 − 40 years many of them
have been regularly timed with high precision es-
pecially in radio observations. Consequently, their
ephemerides (RA, DEC, P, Ṗ, binary orbit param-
eters, pulse arrival time and absolute pulse phase
for a given epoch, pulsar proper motion etc.) are
known with very high accuracy. Indeed, pul-
sar timing has reached the 10−15 fractional level,
which is comparable with the accuracy of atomic
clocks. This is an essential requirement for using
these celestial objects as navigation beacons, as it
enables one to predict the pulse arrival time of a
pulsar for any location in the solar system and be-
yond.

Figure 3: X-ray and radio pulse profiles for the six 3 Principles of Pulsar-Based Navi-
brightest millisecond pulsars. Two full pulse cycles are
shown for clarity. From Becker (2009b). gation
The concept of using pulsars as navigational aids is
X-ray emission, making it possible to study
based on measurements of pulse arrival times and
some of them even with RXTE. For exam-
comparison with predicted arrival times at a given
ple, emission from B1821−24 in the globu-
epoch and reference location. A typical chain
lar cluster M28 is detected by RXTE up to
for detecting, e.g., radio signals from a rotation-
≈ 20 keV, albeit with limited photon statistics.
powered pulsar is shown in Figure 4. An impor-
For the remaining millisecond pulsars the X-
tant step in this measurement is the barycenter cor-
ray emission is found to be much softer, and
rection of the observed photon arrival times. The
pulse profiles are more sinusoidal. Their typ-
pulsar ephemerides along with the position and ve-
ical fraction of pulsed X-ray photons is be-
locity of the observer are parameters of this correc-
tween 30 and 60%. tion. Using a spacecraft position that deviates from
the true position during the observation results in a
Some rotation-powered pulsars have shown phase shift of the pulse peak (or equivalently in a

6
 Telescope

Pulsar

Receiver

TIME
TOA de-dispersion
barycenter correction
coherent folding

Reference clock
Figure 5: Iterative determination of position and ve-
locity by a pulsar-based navigation system.

Figure 4: Typical pulsar detection chain. The pulsar

beams sweep across the radio antenna. Radio signals pulse phase histogram (4) representing the tempo-
are recorded and analyzed in order to produce a mean ral emission characteristics and timing signature of
pulse profile. The data processing comprises a removal the pulsar. This pulse profile, which is continu-
of dispersion effects caused by the interstellar medium
ously improving in significance during an observa-
(“de-dispersion”), correction for the position and proper
motion of the observatory (“barycenter correction”) and tion, is permanently correlated with a pulse profile
coherent folding of many pulses. The time of arrival template in order to increase the accuracy of the
(TOA) of the pulse peak is measured against a reference absolute pulse-phase measurement (5), or equiva-
clock. lently, pulse arrival time (TOA). From the pulsar
ephemeris that includes the information of the ab-
solute pulse phase for a given epoch, the phase dif-
difference in the pulse arrival time). Therefore, the ference ∆φ between the measured and predicted
position and velocity of the spacecraft can be ad- pulse phase can be determined (cf. Figure 6). In
justed in an iterative process until the pulse arrival this scheme, a phase shift (6) with respect to the
time matches with the expected one. The corre- absolute pulse phase corresponds to a range differ-
sponding iteration chain is shown in Figure 5. ence ∆x = cP(∆φ + n) along the line of sight to-
An initial assumption of position and velocity ward the observed pulsar. Here, c is the speed of
is given by the planned orbit parameters of the light, P the pulse period, ∆φ the phase shift and
spacecraft (1). The iteration starts with a pulsar n = 0, ±1, ±2, . . . an integer that takes into account
observation, during which the arrival times of in- the periodicity of the observed pulses. If the phase
dividual photons are recorded (2). The photon shift is non-zero, the position and velocity of the
arrival times have to be corrected for the proper spacecraft needs to be corrected accordingly and
motion of the spacecraft by transforming the ar- the next iteration step is taken (7). If the phase
rival times (3) to an inertial reference location; e.g., shift is zero, or falls below a certain threshold, the
the solar system barycenter (SSB). This correction position and velocity used during the barycenter
requires knowledge of the (assumed or deduced) correction was correct (8) and corresponds to the
spacecraft position and velocity as input parame- actual orbit of the spacecraft.
ters. The barycenter corrected photon arrival times A three-dimensional position fix can be derived
allow then the construction of a pulse profile or from observations of at least three different pul-

7
 SSB
Amplitude (arbitrary units)

Pulsar 1 Pulsar 1

Pulsar 2

Spacecraft

(barycentred) Pu
lsa
r 3

r 4
Amplitude (arbitrary units)

lsa
Pu
Figure 7: Solving the ambiguity problem by observing
four pulsars (drawn in two dimensions). The arrows
point along the pulsar’s lines-of-sight. Straight lines
represent planes of constant pulse phase; black dots in-
Figure 6: Measuring the phase difference between the dicate intersections of planes.
expected and measured pulse peak at an inertial refer-
ence location; e.g., the solar system barycenter (SSB).
The top profile shows the main pulse peak location 4 Radio Antenna for Pulsar-Based
as expected at the SSB. The bottom profile is the one
which has been measured at the spacecraft and trans-
Navigation
formed to the SSB by assuming the spacecraft position
and velocity during the observation. If the position and Pulsars emit broadband electromagnetic radiation.
velocity assumption was wrong, a phase shift ∆φ is ob- Therefore, the observing device of a pulsar-based
served.
navigator can be optimized for the waveband ac-
cording to the highest number of bits per tele-
sars (cf. Figure 7). If on-board clock calibration scope collecting area, power consumption, naviga-
is necessary, the observation of a fourth pulsar is tor weight, cost and compactness. Especially for
required. the question of navigator compactness it is impor-
Since the position of the spacecraft is deduced tant to estimate what size a radio antenna would
from the phase (or pulse arrival time) of a peri- have to have in order to detect the emission from
odic signal, ambiguous solutions may occur. This pulsars in a reasonable integration time. In order to
problem can be solved by constraining the domain estimate this we assume pulsar parameters that are
of possible solutions to a finite volume around typical for millisecond pulsars. As the radio flux
an initial assumed position (Bernhardt et al., 2010, from pulsars shows a ν −1.5 dependence, observa-
2011), or by observing additional pulsars as illus- tions at lower frequencies seem to be preferred, but
trated in Figure 7. scintillation and scattering effects are stronger at

8
lower frequency. For a navigation system operat- very practical for navigation purposes. A naviga-
ing in the radio band of the electromagnetic spec- tor would have to observe several pulsars, either at
trum, the L-band at 21 cm might therefore be best the same time or in series. The pulsars must be
suited. located in different sky regions in order to get an
For a pulsar detection we require a signal-to- accurate navigation result in the x, y and z direc-
noise ratio of S/N = 10, a minimum integration tion. This, however, means that one either rotates
time of tint = 3600 s and assume a frequency band- the parabolic antenna or, alternatively, the whole
width of ∆ν = 100 MHz. For the receiver noise satellite to get the pulsar signals into the antenna
temperature we take Trec = 100 K. A lower temper- focus. Rotating the satellite, though, would mean
ature would require active, e.g., cryogenic cooling, that the communication antenna will not point to
which would increase cost, weight, and power con- earth any more, which is undesirable. On earth
sumption of a navigator. Furthermore, active cool- satellites and space missions to the inner solar sys-
ing would severely limit the lifetime of the navi- tem, power is usually generated by solar panels.
gator due to consumables like helium. For the sky Rotating the satellite would then also mean to bring
temperature we take Tsky = 5 K and for the tele- the solar panels out of optimal alignment with the
scope efficiency ε = 0.5. If Aant is the geometri- sun, which is another counter argument for using
cal antenna area, the effective antenna area com- a parabolic antenna, not to mention the effects of
putes as Aeff = ε Aant . For the period of the pul- shadowing of the solar panels.
sar we assume P = 10 ms and for the pulse width It thus seems more reasonable to use dipole-
W = 2 ms. For the average flux density we adopt array antennas for the pulsar observations. Single
∆S = 10 mJy. Using the canonical sensitivity equa- dipole antennas organized, e.g., as antenna patches
tion (Lorimer & Kramer, 2005) which corresponds could be used to build a larger phased array an-
to the radiometer equation applied to pulsar obser- tenna. Such a phased array would still be large and
vations heavy, though. Depending on the frequency, 104 –
2k (Trec + Tsky )
r
W 105 single patches are required. There have been
∆Smin = no phased-array antennas of that size been build
ε Aant n p tint ∆ν P −W
p
for use in space so far, although smaller prototypes
and converting it to the geometrical antenna area, exist (Datashvili et al., 2011). From them one may
including the S/N requirement, we get: estimate the weight of such antenna arrays. As-
r suming an antenna thickness of 1 cm and an aver-
S 2k (Trec + Tsky ) W
Aant = √ aged density of the antenna material of 0.1 g/cm3
N ε ∆S 2tint ∆ν P −W still yields an antenna weight of 170 kg for the
Here we have also assumed that both polarizations 170 m2 patched antenna array. The signals from
are averaged (n p = 2). From this we compute an the single dipole antennas have to be correlated in
antenna area of Aant ≈ 342 m2 for the parameters phase to each other, which means that all patches
specified above. For a parabolic antenna it would have to be connected to each other by a wired
mean a radius of about Aant /π ≈ 10.5 m. In-
p
mesh and phase correlators. If this phase corre-
creasing the integration time to 4 hours, we find lation and the real-time coherent correction for the
an area of ≈ 171 m2 , which corresponds to a ra- pulse broadening by interstellar dispersion is done
dius of ≈ 7.3 m. For comparison, the radius of by software, it requires a computer with a Terra-
the communication antenna used on Cassini and flop GPU of about 500 W power consumption. A
Voyager is 2 m. It may depend on the satellite clear advantage of a phased antenna array would be
platform what size and antenna weight is accept- that it allows to observe different pulsars located in
able, but a parabolic antenna does not seem to be different sky regions at the same time. That means

9
of course that such an antenna can be smaller if by a pulsar-based navigation system in view of the
the same S/N ratio is to be achieved for a num- still limited information we have today on pulse
ber Nsources of sources within a given time. With profiles and absolute pulse phases in the X-ray
a single-dish antenna one would have to increase band. To overcome the limitations introduced by
1/4
its diameter by Nsources if these had to be observed improper fitting functions, undefined pivot points
within the same given time interval. of pulse peaks and phase shifts by an unmodeled
energy dependence in the pulsed signal, we con-
structed pulse profile templates for all pulsars for
5 Using X-ray Signals from Pul- which pulsed X-ray emission is detected. Where
sars for Spacecraft Navigation supported by photon statistics, templates were con-
structed for various energy ranges. These energy
The increasing sensitivity of the X-ray observa- ranges were chosen in order to optimize the S/N
tories ROSAT, RXTE, XMM-Newton and Chan- ratio of the pulsed signal while sampling as much
dra allowed for the first time to explore in de- as possible of the energy dependence of the X-ray
tail the X-ray emission properties of a larger pulses.
sample of rotation-powered pulsars. The dis- In the literature various authors applied differ-
covery of pulsed X-ray emission from millisec- ent analysis methods and often used different def-
ond pulsars (Becker & Trümper, 1993), the de- initions for pulsed fraction and pulse peak pivot
termination of the X-ray efficiency of rotation- points. Reanalyzing all data from X-ray pulsars
powered pulsars (Becker & Trümper, 1997) as available in the public XMM-Newton, Chandra
well as discoveries of X-ray emission from and RXTE data archives was therefore a require-
various pulsars (e.g., Becker & Trümper, 1999; ment to reduce systematic uncertainties that would
Becker et al., 2005, 2006) and their detailed spa- have been introduced otherwise. The result is a
tial, spectral and timing studies are just a few database containing the energy dependent X-ray
of many accomplishments worth mentioning in pulse profiles, templates and relevant timing and
this context (e.g., Pavlov et al., 2001; Kuiper et al., spectral properties of all X-ray pulsars that have
2002; Weisskopf et al., 2004; De Luca et al., 2005; been detected so far (Prinz, 2010; Breithuth, 2012).
Becker, 2009a; Hermsen et al., 2013). With these According to the harmonic content of an X-ray
new results at hand it was only natural to start look- pulse the templates were obtained by fitting the ob-
ing at their applicability to, e.g., spacecraft nav- served pulse profiles by series of Gaussian and si-
igation based on X-ray data from pulsars. The nusoidal functions. The database further includes
prospects of this application are of even further information on the local environment of a pulsar,
interest considering that low-mass X-ray mirrors, i.e., whether it is surrounded by a plerion, super-
which are an important requirement for a realistic nova remnant or whether it is located in a crowded
implementation of such a navigation system, have sky region like a globular cluster. The latter has
been developed for future X-ray observatories. a severe impact on the detectability of the X-ray
Given the observational and systematical limita- pulses as it reduces the S/N ratio of the pulsed
tions mentioned above it was a question of general emission by the DC emission from background
interest, which we found not considered with suffi- sources. This in turn is important for the selection
cient gravity in the literature, whether it would be of the optimal pulsars that emit pulses, e.g., in the
feasible to navigate a spaceship on arbitrary orbits hard band (above ≈ 3 keV) in order to blend away
by observing X-ray pulsars. the softer emission from a supernova remnant or
In order to address the feasibility question we plerion.
first determined the accuracy that can be achieved The pulse profile templates allow us to mea-

10
sure pulse arrival times with high accuracy even the linear dependence on P that millisecond pul-
for sparse photon statistics by using a least-square sars are better suited for navigation than those with
fit of an adequately adjusted template. The error larger rotation periods.
of pulse-arrival-time measurements is dominated The precision of a pulsar-based navigation sys-
by the systematic uncertainty that comes with the tem thus strongly depends on the choice of pulsars
limited temporal resolution of the observed pulse and the accuracy of pulse arrival measurements,
profiles used to construct the templates. The sta- which is subject to the quality of the available tem-
tistical error in fitting a measured profile by a tem- plates, accuracy of the on-board clock and clock
plate was found to be much smaller in all cases. calibration. As mentioned above, in order to ob-
Assuming that the temporal resolution of the de- tain three-dimensional position information, tim-
tector is not the limiting factor, the temporal reso- ing of at least three different pulsars has to be per-
lution of a pulse profile is given by the widths of formed. The spatial arrangement of these pulsars is
the phase bins used to represent the observed X- another parameter of the achievable accuracy. Our
ray pulse. The bin width, or the number of phase simulations show that the systematic error of po-
bins applied, is a compromise between maximiz- sition determination can be reduced significantly
ing the S/N ratio per phase bin while sampling as by choosing a pulsar triple that is optimal in the
much of the harmonic content as possible. De- sense that the pulsars are nearly perpendicular to
noting the Fourier-power of the i-th harmonic by each other. Since a pulsar might be obscured by
Ri and taking m as the optimal number of har- the sun or a planet and, therefore, its availability
monics deduced from the H-test (De Jager et al., for navigation depends on the current position of
1989), an exact expression for the optimal number the spacecraft, the optimal pulsar triple has to be
of phase bins is given by M = 2.36(∑m 2 2 1/3
i=1 i Ri ) selected from a ranking of possible pulsar combi-
(Becker & Trümper, 1999). This formula com- nations. The following Table 1 represents the rank-
promises between information lost due to binning ing of pulsar combinations, which according to our
(i.e., zero bin width to get all information), and the analysis provide the highest position accuracy via
effect of fluctuations due to finite statistics per bin pulsar navigation (Bernhardt et al. 2010).
(i.e., bin width as large as possible to reduce the
statistical error per bin). The total error (bias plus Rank Pulsar 3-Combination
variance) is minimized at a bin width of 1/M. We 1 B1937+21 B1821−24 J0030+0451
applied this in the reanalysis of pulse profiles in 2 B1937+21 B1821−24 J1023+0038
our database. Pulsed fractions were computed by 3 B1821−24 J0030+0451 J0437−4715
4 B1937+21 J1023+0038 J0218+4232
applying a bootstrap method (Becker & Trümper,
5 B1821−24 J1023+0038 J0437−4715
1999), which again leads to results that are not bi- 6 B1937+21 J0030+0451 J0218+4232
ased by the observers “taste” on where to assume 7 B1937+21 B1821−24 J0437−4715
the DC level in a profile. 8 B1937+21 J0218+4232 J0437−4715
The minimal systematic phase uncertainty for 9 B1821−24 J0218+4232 J0437−4715
10 J1023+0038 J0218+4232 J0437−4715
the pulse profile templates in our database is of the
order of 0.001 (Bernhardt et al., 2010). This un-
Table 1: Ranking of pulsar 3-combinations according
certainty multiplied by the rotation period P of the to the position accuracy achievable when using them
pulsar yields the uncertainty in pulse arrival time in a navigation system based onf X-ray pulsars . All
due to the limited information we have on the exact listed sources are solitary millisecond pulsars except
X-ray pulse profile. Multiplying this in turn by the J0437−4715, which is in a binary.
speed of light yields the spacecraft’s position error
along the line of sight to the pulsar. It is evident by For the pulsars ranked highest in Table 1 we

11
found position errors of about 5 km as a lower limit sult of this ongoing project will be a high-level de-
(cf. Figure 8). The ranking is independent from a sign of a pulsar navigator that accounts for bound-
specific spacecraft orbit, but was obtained under ary conditions (e.g., weight, cost, complexity and
the assumption that the navigation system is capa- power consumption) set by the requirements of a
ble of measuring pulse profiles with the same level specific spacecraft and mission design.
of detail and accuracy as the ones used in the sim- The chart shown in Figure 9 illustrates the work
ulation. Indeed, this is a severe limitation as those logic of our current simulations. Navigator bound-
pulse profiles where obtained by powerful X-ray ary conditions (1) and spacecraft orbit (2) are pre-
observatories like XMM-Newton and Chandra. It defined and constrain the technology parameters
is unlikely, due to weight constraints and power (3) of the navigator’s X-ray detector, mirror system
limitations, that a navigation system will have sim- and on-board electronics. Examples of parame-
ilar capacities in terms of collecting power, tempo- ters that will be analyzed in the simulations are de-
ral resolution and angular resolution. tector technology, temporal and energy resolution,
on-board-clock accuracy and stability, mirror tech-
nology along with collecting power, angular and
spatial resolution, focal length, field of view – just
to mention the most important ones. Given their
X-ray emission properties some pulsars may not
be detectable by the navigator because of its lim-
ited angular resolution and sensitivity. The prop-
erties of the detector and mirror system along with
the trajectory of the virtual navigator thus deter-
mine the set of available pulsars (5), from which
a suitable selection has to be made (6) according
to a predefined ranking of pulsar triples. Using an
X-ray-sky simulator (7), which was developed to
Figure 8: Spacecraft position error as a function of simulate observations of the future X-ray observa-
possible pulsar 3-combinations. The diagram shows tories eROSITA and IXO and which we have mod-
the mean position errors and standard deviations for the ified to include the temporal emission properties
best 30 combinations. From Bernhardt et al. (2010). of X-ray pulsars, we will be able to create X-ray
datasets in FITS-format with temporal, spatial and
An improved accuracy can be achieved by energy information for the pulsars observed by our
means of pulse profile templates of better quality. virtual navigator. The simulated event files have
This, in turn, calls for deeper pulsar observations the same standard FITS-format as those of XMM-
by XMM-Newton or Chandra as long as these ob- Newton and/or Chandra, so that standard software
servatories are still available for the scientific com- can be used to analyze these data. An autonomous
munity. It would be a valuable task worthwhile data reduction will then perform the data analysis
the observing time, especially as it is unclear what and TOA measurement in order to obtain the posi-
missions will follow these great observatories and tion and velocity of the virtual observer (8). Corre-
whether they will provide detectors with sufficient lating the result with the input trajectory (9) yields
temporal resolution and on-board clock accuracy. the accuracy of the simulated measurements for a
We extended our simulations in order to con- given spacecraft orbit (10), and hence the overall
strain the technological parameters of possible performance of the navigator as a function of the
navigation systems based on X-ray pulsars. The re- specified detector and mirror system (11,12).

12
Figure 9: Work flow and logic of the simulations performed for a technology requirement study and demonstrator
high-level design of a pulsar-based navigator.

13
6 X-ray Detector and Mirror bolic surface. This geometry allows for nesting
Technology for Pulsar-Based several concentric mirror shells into each other in
order to enlarge the collecting area and thereby im-
Navigation prove the signal-to-noise ratio. A novel approach
to X-ray optics is the use of pore structures in a
The design of an X-ray telescope suitable for nav-
Wolter-I configuration (Bavdaz et al., 2003, 2010;
igation by X-ray pulsars will be a compromise
Beijersbergen et al., 2004b). X-ray photons that
between angular resolution, collecting area and
enter a pore are focused by reflections on the walls
weight of the system. The currently operating
inside the pore. In contrast to traditional X-ray op-
X-ray observatories XMM-Newton and Chandra
tics with separate mirror shells that are mounted
have huge collecting areas of 0.43 m2 and 0.08 m2
to a support structure, pore optics form a mono-
(at 1 keV), respectively (Friedrich, 2008, and ref-
lithic, self-supporting structure that is lightweight,
erences therein) and, in the case of Chandra, attain
but also very stiff and contains many reflecting sur-
very good angular resolution of less than 1 arcsec-
faces in a compact assembly. Two different types
ond. Their focusing optics and support structures,
of pore optics have been developed, based on sili-
however, are very heavy. To use their mirror tech-
con and glass.
nology would be a show-stopper for a navigation
system.
• Silicon Pore Optics (Collon et al., 2009,
In recent years, ESA and NASA have put
2010; Ackermann et al., 2009) use commer-
tremendous effort into the development of low-
cially available and mass produced silicon
mass X-ray mirrors, which can be used as ba-
wafers (Figure 10a) from the semiconduc-
sic technology for future large X-ray observatories
tor industry. These wafers have a surface
and small planetary exploration missions. Table
roughness that is sufficiently low to meet
2 summarizes the angular resolution and mass of
the requirements of X-ray optics. A chemo-
X-ray mirrors used for XMM-Newton and Chan-
mechanical treatment of a wafer results in a
dra as well as developed for future X-ray missions.
very thin membrane with a highly polished
The light-weighted mirrors are of special interest
surface on one side and thin ribs of very ac-
for an X-ray pulsar-based navigator.
curate height on the other side. Several of
these ribbed plates are elastically bent to the
Angular Mass per effective
geometry of a Wolter-I system, stacked to-
resolution area (at 1 keV)
gether to form the pore structure and finally
Chandra 0.5′′ 18 500 kg/m2
XMM-Newton 14′′ 2300 kg/m2
integrated into mirror modules (Figure 10a).
Silicon Pore Optics 5′′ 200 kg/m2 Silicon Pore Optics are intended to be used on
Glass Micropore Optics 30′′ 25 kg/m2 large X-ray observatories that require a small
mass per collecting area (on the order of 200
Table 2: Comparison of current and future X-ray- kg/m2 ) and angular resolution of about 5 arc-
mirror optics. From Bavdaz et al. (2010). seconds or better.

A typical high-resolution X-ray telescope uses • Glass Micropore Optics

focusing optics based on the Wolter-I design (Beijersbergen et al., 2004a; Collon et al.,
(Wolter, 1952). The incoming X-ray photons are 2007; Wallace et al., 2007) are made from
reflected under small angles of incidence in or- polished glass blocks that are surrounded by
der not to be absorbed and are focused by dou- a cladding glass with a lower melting point.
ble reflection off a parabolic and then a hyper- In order to obtain the high surface quality

14
 which the remaining cladding glass forms the
pore walls (Figure 10b). The Wolter-I ge-
ometry is reproduced by thermally slumping
separate multi-fiber plates. Glass Micropore
Optics are even lighter than Silicon Pore
Optics, but achieve a moderate angular
resolution of about 30 arcseconds. They are
especially interesting for small planetary ex-
ploration missions, but also for X-ray timing
a. missions that require large collecting areas.
The first implementation of Glass Micropore
Optics on a flight program will be in the
Mercury Imaging X-ray Spectrometer on the
ESA/JAXA mission BepiColombo, planned
to launch in 2014 (Fraser et al., 2010).
Today’s detector technology, as in use on XMM-
Newton and Chandra, is not seen to provide a de-
tector design that is useful for a navigation sys-
b.
b. tem based on X-ray pulsars. Readout noise, lim-
ited imaging capability in timing-mode and out-of-
time events invalidate CCD-based X-ray detectors
for application as X-ray-pulsar navigator. Detec-
tors like those on RXTE that need gas for oper-
OPTICAL AXIS

ation are also not suitable, given the limited live

time due to consumables. However, there are novel
and promising detector developments performed in
semiconductor labs for the use in the next genera-
tion of X-ray observatories. Two challenging ex-
c. amples, which are of potential interest for naviga-
tion, are:
Figure 10: Silicon pore optics (a) and glass micro- • Silicon Drift Detectors (SDDs) have only
pore optics (b) represent novel developments for light- limited imaging capability but provide an en-
weighted X-ray mirrors of the next generation of X-ray ergy resolution and are capable of managing
observatories. Both mirror types will be used in Wolter-
high counting rates of more than 10 Crab (1
I configuration (c) to focus X-rays in a double reflec-
tion. Images from Bavdaz et al. (2003, 2004, 2010). Crab count rate ≈ 200 000 cts/s). A detec-
tor based on this technology was proposed
for the High Time Resolution Spectrometer
required for X-ray optics, the blocks are on IXO (Barret et al., 2010). The detector
stretched into small fibers, thereby reducing technology itself has a high technical readi-
the surface roughness. Several of these fibers ness. SDD-modules developed in the Semi-
can be assembled and fused into multi-fiber conductor Lab of the Max Planck Society are
bundles. Etching away the glass fiber cores working already in the APXS (Alpha Par-
leads to the desired micropore structure, in ticle X-ray Spectrometer) on-board NASA’s

15
 Mars Exploration Rovers Spirit and Oppor-
tunity and on the comet lander ROSETTA
(Lechner et al., 2010). Detectors based on an
SDD technology could be of use, e.g., in de-
signing a navigator for a very specific orbit,
for which it is sufficient to navigate accord-
ing to the signals of pulsars that emit their
pulses in the hard X-ray band mostly so that
the missing imaging capability does not cause
any restrictions on the S/N ratio of the pulsed
emission. Figure 11: Mechanical sample of an Active Pixel (here
6-inch wafer-scale) detector. Plotted over one hemi-
• Active Pixel Sensors (APS) are an alterna- sphere is the logical layout of the detector. It consists
tive and perhaps more flexible technology of roughly 1024 × 1024 pixels of 100 × 100 µ m2 size.
(cf. Figure 11), which was the proposed tech- From Lechner et al. (2010).
nology for the Wilde Field Imager on IXO
(Strüder et al., 2010) and the Low-Energy De- Simbol-X had 128 × 128 pixels and an energy con-
tector (LED) on Simbol-X (Lechner, 2009). sumption of ≤ 8 W. Power consumption including
This detector provides images in the energy electronics, filter wheel and temperature control
band 0.1–25 keV, simultaneously with spec- was ca. 250 W. The mass of the focal plane, in-
trally and time resolved photon counting. The cluding shielding and thermal interface, was about
device, which is under development in the 15 kg but could be reduced in a more specific de-
MPE Semiconductor Laboratory, consists of sign of an X-ray-pulsar navigator.
an array of DEPFET (Depleted p-channel
FET) active pixels, which are integrated onto
a common silicon bulk. The DEPFET con- 7 Concluding Remarks
cept unifies the functionalities of both sen-
sor and amplifier in one device. It has a sig- The knowledge of how to use stars, planets and
nal charge storage capability and is read out stellar constellations for navigation was fundamen-
demand. The DEPFET is used as unit cell tal for mankind in discovering new continents and
of Active Pixel Sensors (APSs) with a scal- subduing living space in ancient times. It is fas-
able pixel size from 50 µ m to several mm and cinating to see how history repeats itself in that a
a column-parallel row-by-row readout with a special population of stars may play again a funda-
short signal processing time of ≤ 4 µ sec per mental role in the future of mankind by providing
row. As the pixels are individually address- a reference for navigating their spaceships through
able the DEPFET APS offers flexible readout the Universe (cf. Figure 12).
strategies from standard full-frame mode to
In the paper we have shown that autonomous
user-defined window mode. spacecraft navigating with pulsars is feasible when
The typical weight and power consumption of using either phased-array radio antennas of at least
these detectors can be estimated from the proto- 150 m2 antenna area or compact light-weighted X-
types proposed for IXO and Simbol-X. The IXO ray telescopes and detectors, which are currently
Wide Field Imager with its 17 arcmin field of view developed for the next generation of X-ray obser-
and 1024 × 1024 pixel design had an energy con- vatories.
sumption of ≤ 22 W. The Low Energy Detector on Using the X-ray signals from millisecond pul-

16
Figure 12: Artist’s impression of Rosetta, if it navigated in deep space using pulsar signals. The characteristic
time signatures of pulsars are used as natural navigation beacons to determine the position and velocity of the
spacecraft.

sars we estimated that navigation would be possi- of the receiving system and, therefore, cannot be-
ble with an accuracy of ±5 km in the solar system come arbitrarily large. In addition, irradiation from
and beyond. The error is dominated by the inaccu- the on-board electronics requires an efficient elec-
racy of the pulse profiles templates that were used tromagnetic shielding to prevent signal feedback.
for the pulse peak fittings and pulse-TOA measure- This shielding will further increase the navigator
ments. As those are known with much higher ac- weight in addition to the weight of the antenna.
curacy in the radio band, it is possible to increase The optimal choice of the observing band de-
the accuracy of pulsar navigation down to the me- pends on the boundary conditions given by a spe-
ter scale by using radio signals from pulsars for cific mission. What power consumption and what
navigation. navigator weight might be allowed for may deter-
The disadvantage of radio observations in a nav- mine the choice for a specific wave band.
igator, though, is the large size and mass of the In general, however, it is clear already today that
phased-antenna array. As we saw in § 4, the an- this navigation technique will find its applications
tenna area is inversely proportional to the square in future astronautics. The technique behind it is
root of the integration time; i.e., the same signal very simple and straightforward, and pulsars are
quality can be obtained with a reduced antenna available everywhere in the Galaxy. Today ≈ 2200
size by increasing the observation time. However, pulsars are known. With the next generation of ra-
the observing time is limited by the Allen variance dio observatories, like the SKA, it is expected to

17
detect signals from about 20 000 to 30 000 pulsars M., & Mereghetti, S. (2010). The High Time
(Smits et al., 2009). Resolution Spectrometer (HTRS) aboard the In-
Finally, pulsar-based navigation systems can op- ternational X-ray Observatory (IXO). In Soci-
erate autonomously. This is one of their most ety of Photo-Optical Instrumentation Engineers
important advantages, and is interesting also for (SPIE) Conference Series, volume 7732 of Soci-
current space technologies; e.g., as augmentation ety of Photo-Optical Instrumentation Engineers
of existing GPS/Galileo satellites. Future appli- (SPIE) Conference Series.
cations of this autonomous navigation technique
might be on planetary exploration missions and on Battin, R. H. (1964). Astronautical Guidance.
manned missions to Mars or beyond. McGraw-Hill, New York, USA.

Bavdaz, M., Collon, M., Beijersbergen, M., Wal-

Acknowledgments lace, K., & Wille, E. (2010). X-Ray Pore Optics
Technologies and Their Application in Space
WB acknowledges discussion with David Cham- Telescopes. X-Ray Optics and Instrumentation,
pion (MPIfR), Horst Baier and Ulrich Walter 2010, 1–15.
(TUM). MGB acknowledges support from and
participation in the International Max Planck Re- Bavdaz, M., Peacock, A. J., Lehmann, V., Beijer-
search School on Astrophysics at the Ludwig Max- sbergen, M. W., & Kraft, S. (2003). X-ray Op-
imilians University of Munich, Germany. tics: new technologies at ESA. In Proc. SPIE,
volume 4851 (pp. 421–432).

References Bavdaz, M., Peacock, A. J., Tomaselli, E., Bei-

jersbergen, M. W., Collon, M., Flyckt, S.-O.,
Ackermann, M. D., Collon, M. J., Guenther, R., Fairbend, R., & Boutot, J.-P. (2004). Progress
Partapsing, R., Vacanti, G., Buis, E.-J., Krum- at ESA on high-energy optics technologies.
rey, M., Müller, P., Beijersbergen, M. W., Bav- In O. Citterio and S. L. O’Dell (Ed.), Soci-
daz, M., & Wallace, K. (2009). Performance ety of Photo-Optical Instrumentation Engineers
prediction and measurement of Silicon Pore Op- (SPIE) Conference Series, volume 5168 of Soci-
tics. In Proc. SPIE, volume 7437 (pp. 74371N1– ety of Photo-Optical Instrumentation Engineers
10). (SPIE) Conference Series (pp. 136–147).
Backer, D. C., Kulkarni, S. R., Heiles, C., Davis, Becker, W., Ed. (2009a). Neutron Stars and Pul-
M. M., & Goss, W. M. (1982). A millisecond sars, volume 357 of Astrophysics and Space Sci-
pulsar. Nature, 300, 615–618. ence Library. Springer, Berlin, Germany.
Barret, D., Ravera, L., Bodin, P., Amoros, C., Becker, W. (2009b). X-Ray Emission from Pul-
Boutelier, M., Glorian, J.-M., Godet, O., Ort- sars and Neutron Stars. In W. Becker (Ed.),
tner, G., Lacombe, K., Pons, R., Rambaud, D., Neutron Stars and Pulsars, volume 357 of As-
Ramon, P., Ramchoun, S., Biffi, J.-M., Bela- trophysics and Space Science Library (pp. 91–
sic, M., Clédassou, R., Faye, D., Pouilloux, B., 140).: Springer, Berlin, Germany.
Motch, C., Michel, L., Lechner, P. H., Niculae,
A., Strueder, L. W., Distratis, G., Kendziorra, E., Becker, W., Jessner, A., Kramer, M., Testa, V., &
Santangelo, A., Tenzer, C., Wende, H., Wilms, Howaldt, C. (2005). A Multiwavelength Study
J., Kreykenbohm, I., Schmid, C., Paltani, S., of PSR B0628−28: The First Overluminous
Cadoux, F., Fiorini, C., Bombelli, L., Méndez, Rotation-powered Pulsar? ApJ, 633, 367–376.

18
Becker, W., Kramer, M., Jessner, A., Taam, R. E., Bhattacharya, D. & van den Heuvel, E. P. J. (1991).
Jia, J. J., Cheng, K. S., Mignani, R., Pellizzoni, Formation and evolution of binary and millisec-
A., de Luca, A., Słowikowska, A., & Caraveo, ond radio pulsars. Phys. Rep., 203, 1–124.
P. A. (2006). A Multiwavelength Study of the
Pulsar PSR B1929+10 and Its X-Ray Trail. ApJ, Blandford, R. & Teukolsky, S. A. (1976). Arrival-
645, 1421–1435. time analysis for a pulsar in a binary system.
ApJ, 205, 580–591.
Becker, W. & Trümper, J. (1993). Detection of
pulsed X-rays from the binary millisecond pul- Breithuth, F. M. (2012). Erstellung einer Pulsar-
sar J0437−4715. Nature, 365, 528–530. datenbank und ihre Anwendung für die Sim-
ulation eines pulsarbasierten Navigationssys-
Becker, W. & Trümper, J. (1997). The X-ray lumi- tems. Master’s thesis, Ludwig-Maximilians-
nosity of rotation-powered neutron stars. A&A, Universität München. (In German).
326, 682–691.
Chester, T. J. & Butman, S. A. (1981). Nav-
Becker, W. & Trümper, J. (1999). The X-ray emis- igation Using X-Ray Pulsars. NASA Tech.
sion properties of millisecond pulsars. A&A, Rep. 81N27129, Jet Propulsion Laboratory,
341, 803–817. Pasadena, CA, USA.

Beijersbergen, M., Bavdaz, M., Buis, E. J., & Collon, M. J., Beijersbergen, M. W., Wallace, K.,
Lumb, D. H. (2004a). Micro-pore X-ray optics Bavdaz, M., Fairbend, R., Séguy, J., Schyns, E.,
developments and application to an X-ray tim- Krumrey, M., & Freyberg, M. (2007). X-ray
ing mission. In Proc. SPIE, volume 5488 (pp. imaging glass micro-pore optics. In Proc. SPIE,
468–474). volume 6688 (pp. 668812/1–13).

Beijersbergen, M., Kraft, S., Bavdaz, M., Lumb, Collon, M. J., Guenther, R., Ackermann, M., Par-
D., Guenther, R., Collon, M., Mieremet, A., tapsing, R., Kelly, C., Beijersbergen, M. W.,
Fairbend, R., & Peacock, A. (2004b). Develop- Bavdaz, M., Wallace, K., Olde Riekerink, M.,
ment of x-ray pore optics: novel high-resolution Mueller, P., & Krumrey, M. (2009). Stacking
silicon millipore optics for XEUS and ultralow of Silicon Pore Optics for IXO. In Proc. SPIE,
mass glass micropore optics for imaging and volume 7437 (pp. 74371A1–7).
timing. In Proc. SPIE, volume 5539 (pp. 104–
115). Collon, M. J., Günther, R., Ackermann, M., Par-
tapsing, R., Vacanti, G., Beijersbergen, M. W.,
Bernhardt, M. G., Becker, W., Prinz, T., Breithuth, Bavdaz, M., Wille, E., Wallace, K., Olde Riek-
F. M., & Walter, U. (2011). Autonomous Space- erink, M., Lansdorp, B., de Vrede, L., van
craft Navigation Based on Pulsar Timing Infor- Baren, C., Müller, P., Krumrey, M., & Freyberg,
mation. In 2nd International Conference on M. (2010). Silicon Pore X-ray Optics for IXO.
Space Technology (pp. 1–4). In Proc. SPIE, volume 7732 (pp. 77321F1–9).

Bernhardt, M. G., Prinz, T., Becker, W., & Walter, Datashvili, L., Baier, H., Kuhn, T., Langer, H.,
U. (2010). Timing X-ray Pulsars with Applica- Apenberg, S., & Wei, B. (2011). A Large De-
tion to Spacecraft Navigation. In High Time Res- ployable Space Array Antenna: Technology and
olution Astrophysics IV, PoS(HTRA-IV)050 (pp. Functionality Demonstrator. In Proc. of ESA An-
1–5). tenna Workshop.

19
De Jager, O. C., Raubenheimer, B. C., & Ghosh, P. (2007). Rotation and Accretion Powered
Swanepoel, J. W. H. (1989). A poweful test for Pulsars. World Scientific Publishing, New Jer-
weak periodic signals with unknown light curve sey, USA.
shape in sparse data. A&A, 221, 180–190.
Hermsen, W., Hessels, J. W. T., Kuiper, L., van
De Luca, A., Caraveo, P. A., Mereghetti, S., Ne- Leeuwen, J., Mitra, D., de Plaa, J., Rankin,
groni, M., & Bignami, G. F. (2005). On the Polar J. M., Stappers, B. W., Wright, G. A. E., Basu,
Caps of the Three Musketeers. ApJ, 623, 1051– R., Alexov, A., Coenen, T., Grießmeier, J.-
1069. M., Hassall, T. E., Karastergiou, A., Keane,
Downs, G. S. (1974). Interplanetary Navigation E., Kondratiev, V. I., Kramer, M., Kuniyoshi,
Using Pulsating Radio Sources. NASA Tech. M., Noutsos, A., Serylak, M., Pilia, M., Sobey,
Rep. 74N34150 (JPL Tech. Rep. 32-1594), Jet C., Weltevrede, P., Zagkouris, K., Asgekar, A.,
Propulsion Laboratory, Pasadena, CA, USA. Avruch, I. M., Batejat, F., Bell, M. E., Bell,
M. R., Bentum, M. J., Bernardi, G., Best, P.,
Duncan, R. C. & Thompson, C. (1992). Forma- Bı̂rzan, L., Bonafede, A., Breitling, F., Brod-
tion of very strongly magnetized neutron stars erick, J., Brüggen, M., Butcher, H. R., Ciardi,
– Implications for gamma-ray bursts. ApJ, 392, B., Duscha, S., Eislöffel, J., Falcke, H., Fender,
L9–L13. R., Ferrari, C., Frieswijk, W., Garrett, M. A.,
de Gasperin, F., de Geus, E., Gunst, A. W.,
Espinoza, C. M., Lyne, A. G., Stappers, B. W., & Heald, G., Hoeft, M., Horneffer, A., Iacobelli,
Kramer, M. (2011). A study of 315 glitches in M., Kuper, G., Maat, P., Macario, G., Markoff,
the rotation of 102 pulsars. MNRAS, 414, 1679– S., McKean, J. P., Mevius, M., Miller-Jones,
1704. J. C. A., Morganti, R., Munk, H., Orrú, E., Paas,
Fraser, G. W., Carpenter, J. D., Rothery, D. A., H., Pandey-Pommier, M., Pandey, V. N., Pizzo,
Pearson, J. F., Martindale, A., Huovelin, J., R., Polatidis, A. G., Rawlings, S., Reich, W.,
Treis, J., Anand, M., Anttila, M., Ashcroft, M., Röttgering, H., Scaife, A. M. M., Schoenmak-
Benkoff, J., Bland, P., Bowyer, A., Bradley, ers, A., Shulevski, A., Sluman, J., Steinmetz,
A., Bridges, J., Brown, C., Bulloch, C., Bunce, M., Tagger, M., Tang, Y., Tasse, C., ter Veen, S.,
E. J., Christensen, U., Evans, M., Fairbend, R., Vermeulen, R., van de Brink, R. H., van Weeren,
Feasey, M., Giannini, F., Hermann, S., Hesse, R. J., Wijers, R. A. M. J., Wise, M. W., Wuck-
M., Hilchenbach, M., Jorden, T., Joy, K., Kaip- nitz, O., Yatawatta, S., & Zarka, P. (2013). Syn-
iainen, M., Kitchingman, I., Lechner, P., Lutz, chronous X-ray and Radio Mode Switches: A
G., Malkki, A., Muinonen, K., Näränen, J., Rapid Global Transformation of the Pulsar Mag-
Portin, P., Prydderch, M., Juan, J. S., Sclater, netosphere. Science, 339, 436–.
E., Schyns, E., Stevenson, T. J., Strüder, L.,
Syrjasuo, M., Talboys, D., Thomas, P., Whit- Hewish, A., Bell, S. J., Pilkington, J. D. H., Scott,
ford, C., & Whitehead, S. (2010). The mercury P. F., & Collins, R. A. (1968). Observation of
imaging X-ray spectrometer (MIXS) on bepi- a Rapidly Pulsating Radio Source. Nature, 217,
colombo. Planet. Space Sci., 58(1-2), 79–95. 709–713.

Friedrich, P. (2008). Wolter Optics. In J. E. James, N., Abello, R., Lanucara, M., Mercolino,
Trümper & G. Hasinger (Eds.), The Universe M., & Maddè, R. (2009). Implementation of an
in X-Rays, Astronomy and Astrophysics Library ESA delta-DOR capability. Acta Astronautica,
(pp. 41–50).: Springer, Berlin, Germany. 64(11-12), 1041–1049.

20
Kramer, M., Xilouris, K. M., Lorimer, D. R., Matsakis, D. N., Taylor, J. H., & Eubanks, T. M.
Doroshenko, O., Jessner, A., Wielebinski, R., (1997). A statistic for describing pulsar and
Wolszczan, A., & Camilo, F. (1998). The Char- clock stabilities. A&A, 326, 924–928.
acteristics of Millisecond Pulsar Emission. I.
Spectra, Pulse Shapes, and the Beaming Frac- Pavlov, G. G., Zavlin, V. E., Sanwal, D., Burwitz,
tion. ApJ, 501, 270. V., & Garmire, G. P. (2001). The X-Ray Spec-
trum of the Vela Pulsar Resolved with the Chan-
Kuiper, L., Hermsen, W., Verbunt, F., Ord, S., dra X-Ray Observatory. ApJ, 552, L129–L133.
Stairs, I., & Lyne, A. (2002). High-Resolution
Spatial and Timing Observations of Millisecond Prinz, T. (2010). Zeitanalyse der Röntgenemission
Pulsar PSR J0218+4232 with Chandra. ApJ, von Pulsaren und deren spezielle Anwen-
577, 917–922. dung für die Navigation von Raumfahrzeu-
gen. Diploma thesis, Ludwig-Maximilians-
Lechner, P. (2009). The Simbol-X Low Energy Universität München. (In German).
Detector. In J. Rodriguez and P. Ferrando
(Ed.), American Institute of Physics Conference Riedel, J. E., Bhaskaran, S., Desai, S., Han,
Series, volume 1126 of American Institute of D., Kennedy, B., Null, G. W., Synnott, S. P.,
Physics Conference Series (pp. 21–24). Wang, T. C., Werner, R. A., & Zamani,
E. B. (2000). Autonomous Optical Naviga-
Lechner, P., Amoros, C., Barret, D., Bodin,
tion (AutoNav) DS1 Technology Validation Re-
P., Boutelier, M., Eckhardt, R., Fiorini, C.,
port. Deep Space 1 technology validation re-
Kendziorra, E., Lacombe, K., Niculae, A.,
ports (Rep. A01-26126 06-12), Jet Propulsion
Pouilloux, B., Pons, R., Rambaud, D., Ravera,
Laboratory, Pasadena, CA, USA.
L., Schmid, C., Soltau, H., Strüder, L., Tenzer,
C., & Wilms, J. (2010). The silicon drift detector Smits, R., Kramer, M., Stappers, B., Lorimer,
for the IXO high-time resolution spectrometer. D. R., Cordes, J., & Faulkner, A. (2009). Pulsar
In Society of Photo-Optical Instrumentation En- searches and timing with the square kilometre
gineers (SPIE) Conference Series, volume 7742 array. A&A, 493, 1161–1170.
of Society of Photo-Optical Instrumentation En-
gineers (SPIE) Conference Series. Strüder, L., Aschauer, F., Bautz, M., Bombelli,
L., Burrows, D., Fiorini, C., Fraser, G., Her-
Lorimer, D. R. & Kramer, M. (2005). Handbook
rmann, S., Kendziorra, E., Kuster, M., Lauf, T.,
of Pulsar Astronomy, volume 4 of Cambridge
Lechner, P., Lutz, G., Majewski, P., Meuris, A.,
observing handbooks for research astronomers.
Porro, M., Reiffers, J., Richter, R., Santangelo,
Cambridge University Press, Cambridge, UK.
A., Soltau, H., Stefanescu, A., Tenzer, C., Treis,
Maddè, R., Morley, T., Abelló, R., Lanucara, M., J., Tsunemi, H., de Vita, G., & Wilms, J. (2010).
Mercolino, M., Sessler, G., & de Vicente, J. The wide-field imager for IXO: status and future
(2006). Delta-DOR – a new technique for ESA’s activities. In Society of Photo-Optical Instru-
Deep Space Navigation. ESA Bulletin, 128, 68– mentation Engineers (SPIE) Conference Series,
74. volume 7732 of Society of Photo-Optical Instru-
mentation Engineers (SPIE) Conference Series.
Manchester, R. N., Hobbs, G. B., Teoh, A., &
Hobbs, M. (2005). The Australia Telescope Na- Taylor, Jr., J. H. (1991). Millisecond pulsars: Na-
tional Facility Pulsar Catalogue. AJ, 129, 1993– ture’s most stable clocks. IEEE Proceedings, 79,
2006. 1054–1062.

21
Trümper, J. E., Dennerl, K., Kylafis, N. D., Ertan,
Ü., & Zezas, A. (2013). An Accretion Model
for the Anomalous X-Ray Pulsar 4U 0142+61.
ApJ, 764, 49.

Trümper, J. E., Zezas, A., Ertan, Ü., & Ky-

lafis, N. D. (2010). The energy spectrum of
anomalous X-ray pulsars and soft gamma-ray
repeaters. A&A, 518, A46+.

Wallace, K., Collon, M. J., Beijersbergen, M. W.,

Oemrawsingh, S., Bavdaz, M., & Schyns, E.
(2007). Breadboard micro-pore optic develop-
ment for x-ray imaging. In Proc. SPIE, volume
6688 (pp. 66881C1–10).

Weisskopf, M. C., O’Dell, S. L., Paerels, F., El-

sner, R. F., Becker, W., Tennant, A. F., & Swartz,
D. A. (2004). Chandra Phase-Resolved X-Ray
Spectroscopy of the Crab Pulsar. ApJ, 601,
1050–1057.

Wolter, H. (1952). Spiegelsysteme streifenden

Einfalls als abbildende Optiken für
Röntgenstrahlen. Ann. Phys., 445, 94–114. (In
German).

Wood, K. S. (1993). Navigation studies utilizing

the NRL-801 experiment and the ARGOS satel-
lite. In Proc. SPIE, volume 1940 (pp. 105–116).

Yu, M., Manchester, R. N., Hobbs, G., Johnston,

S., Kaspi, V. M., Keith, M., Lyne, A. G., Qiao,
G. J., Ravi, V., Sarkissian, J. M., Shannon, R., &
Xu, R. X. (2013). Detection of 107 glitches in
36 southern pulsars. MNRAS, 429, 688–724.

22