Journal of Global Positioning Systems (2004)
Vol. 3, No. 1-2: 308-321
Would a GNSS need a backup?
Walter Blanchard
Royal Institut e of Navigation, Trundle Tower Hill, Dorking, Surrey, UK
e-mail: [e-mail] Tel: ++44 1306 884539
Received: 15 Nov 2004 / Accepted: 3 Feb 2005
Abstract. No navigator likes to be totally dependent on
only one navaid – it is an article of faith for many that
there should always be a backup system. Several systems
have been put forward as possible backups for a GNSS
but they seem to have originated more in a generalised
feeling that there ought to be one rather than a
dispassionate examination of what is involved.
GPS/Galileo are radical departures from any previous
concepts of radio navigation aids and a full-blown GNSS
is an even more radical proposal. There is a good deal
more involved than simply engineering and technical
matters. There are the questions of who controls them;
what the customer interface is; who certifies them for use
in safety-related situations; and what legal recourse there
is. On the answers to these questions depends whether a
backup is needed and if so what form it should take. It is
found in this paper that for many non-critical users there
is no need for a backup, and that others who may be
involved in safety-critical situations already have a
backup in the form of their current systems. It is also
found that in fact it may be extremely difficult to
compose a GNSS in the form it is generally given; that is,
a combination of GPS, Galileo and perhaps Glonass. Th e
problem lies not in on the engine ering side, but in matters
of legality and the sovereignty of individual nations. For
these reasons it is concluded that the development or
implementation of a new system purely to act as a backup
for a GNSS is not ne ce ss ary.
Key words: Satellites, Navigation, GNSS.
1 Introduction
Most people have a slightly uneasy feeling that if they
rely a great deal on something there ought to be a backup
for it, or at least some sort of fall-back provision. For
example, cars; if our car breaks down unexpectedly we
may have a major problem. No-one wants to be stuck
several hundred miles from home. Although fortunately
this is now a rare occurrence with modern cars if it
happens a mobile telephone will bring a breakdown truck
to the rescue. This combination of basic reliability and
the availability of rescue services means that very few
drivers are to be found towing a spare car behind them
“just in case” (except perhaps for large RV owners!).
What was really a similar consideration occurred in the
aviation world some time ago when the wisdom of
allowing twin-engine commercial airliners to transit
oceans was much debated. It was the legacy of
comparatively unreliable piston engines that caused the
debate; the fact that the modern jet engine is far more
reliable took some time to establish and now most trans-
oceanic aircraft are twin-engined.
The same sort of thing is occurring in the navigation
business now we have now come to use GPS so much.
There is a popular feeling there ought to be a “backup” in
case it is “switched off” or somebody jams it. This feeling
is often reinforced by previous experience with radio-
based systems when a transmitter failed or a receiver
packed up, and of course there is alw ays the wise old saw
of never relying totally on one system. But, looking
ahead, how much of this will apply to GNSS, and, if it
does, what type of backup should there be? The rather
vague feeling that “there ought to be a back-up” is not
much of a guide. Many factors are invo lved ranging fro m
the effect withdrawal of GNSS would have on a users’
operations and how a back-up would enable him to
continue them, to the cost of installing back-up
infrastructure and, not least, the cost to individual users of
installing back - u p us er equipment.
In any case, just what IS a back-up? How do you define
the point at which “loss of GNSS services” becomes so
severe that a backup is needed? Should it be able to
completely take over every function of a GNSS; most of
them, or only the most essential? If it can do the same job
as a GNSS, why hold it in reserve? It would be expensive
and if not used except on very rare occasions would be a

Blanchard: Would a GNSS need a backup? 309
financial disaster. Does every user actually need a back-
up? Do different users need different back-ups? Who
would pay for it, run it, control it?
At the moment we have no direct experience with a
GNSS; all we have had so far is GPS, with a few people
using Glonass as well. Perhaps our experience with GPS
might provide a guide, but we must be clear about the
distinction. GPS is owned, operated and controlled
entirely by one country, the United States. A GNSS, by
definition, will consist of a combination of several
distinct systems possibly in multiple ownership but
essentially under consensus-control authority.
Fig. 1 GPS
Fig. 2 Galileo .

310 Journal of Global Positioning Systems
2. THE GPS SITUATION; NON -MANDATORY
USERS.
Fig. 3 GNSS Market (from a diagram by EC GJU).
This diagram shows that by far the great majority of users
of a GNSS will in the category of “non-mandatory” users
– those who are not compelled by law to carry a satnav
system. They might be termed “voluntary” users.
Fig. 4 “Voluntary” user s.
So let’s start by looking at how GPS is used by them. The
category includes the millions of “pleasure” sailors and
aviators; the even more numerous car-drivers who found
a GPS in their new car; and all those others using it on
what might be called a “hobby” basis.
For most of them, GPS is a luxury and many do not even
bother to learn how to use it properly. Its loss would be
inconsequential and they would just fish out the old map
every car has under the seat. Commercial vehicles use it
rather more seriously to find addresses and alternative
routes but here again no commercial driver is without his
A-to-Z and a good map. Discretional vehicle-tracking
systems (those not required by law) would undoubtedly
miss it but not all use GPS. If any of these users really
felt it was essential to have it at all times they would
either install a system not based on GPS, or make
arrangements for their own back-up. Datatrak and other
non-satellite systems are well-established and available
for those who do not want to trust GPS.
As regards light aviation, the standard GA aircraft radio-
navigation fit is still VOR, maybe a DME, an ADF and a
transponder, which will be there whether or not GPS is
fitted. There is no proposal to withdraw any of these in
the foreseeable future. Although GPS is considered an
excellent navaid they will always be there to fall back on
should GPS fail.
Small-boat sailors invariably use GPS now, but before
GPS they seemed to manage quite well using traditional
systems – DR, perhaps a sextant, radio bearings, etc. The
only tricky situation that might happen is if they set out
with GPS and found halfway through the trip it
disappeared. But exactly the same thing would happen if
their GPS batteries went flat and presumably they always
have this possibility in mind.
In the ultimate, if any of this category of user cannot
make a trip without GPS, they are in the happy position
of being able to cancel it, unlike commercial users. The
expense of installing another system purely as a GPS
backup is simply not justified for these non-mandatory
users.
3 THE GPS SITUATION; COMMERCIAL USERS.
But what about tho se for whom GPS is important in their
commercial activities? What about professional aviators
and sailors; land surveyors; offshore surveyors; those
using GPS timing for keeping cellphones working, etc?
Fig. 5 Commercial Users.

Blanchard: Would a GNSS need a backup? 311
We can dismiss commercial aviation immediately, for
much the same reasons as light aviation. VOR/DME, ILS
and INS are mandatory as they have been for many years,
and no aircraft can be certificated for IFR flight or
allowed to land at major airports without them in some
combination or other. Fig. 6 illustrates the profusion of
such aids available in a major terminal area.
Fig. 6 London FIR Aeronautical Chart.
Although GPS is often also fitted it is only as a
supplementary aid which means prima facie the owner
must demonstrate he can navigate without it, in other
words, that he has an adequate fit of the primary systems.
There is therefore no need for the commercial aviator to
have a back-up for GPS, he does not rely on it. Nor is he
likely to do so for many years – there is no proposal to
allow the use of unaided GPS as a sole system, and ICAO
has stated no such inten tion. On e very go od reason is that
it has been adequately demonstrated that unaided GPS
does not have sufficient availability for sole use and it
will certainly need the addition of another system such as
Glonass or Galileo, thus turning it into a GNSS, before it
can be considered. GPS augmentation systems such as
WAAS/LAAS and EGNOS have their own problems and
it may be some time before they are worked out. The
professional aviator will need so me very hard convincing
before he will be happy to ditch his well-established
VOR/DME in favour of a satnav alternative and it is only
then that talk of a backup will become relevant. Even
then, the more likely course is that having already
amortised their conventional VOR/DME, etc

312 Journal of Global Positioning Systems
installations, airlines (and civil aviation administrations)
will insist on them remaining as the major reversionary
system for GNSS. Talk of a different backup, requiring
completely new equipment and installation costs, will be
dismissed instantly. It is often forgotten that the real costs
of putting a new piece of kit into an aircraft lie in its
installation and integration costs, the costs of setting up
workshop support, retraining engineers, and so on, not
just the bare cost of the boxes. Quite apart from these
technical arguments, the problem of legislation in
countries having no control over GPS seems insoluble at
present. For these reasons, arguments using commercial
aviation as justification for a GPS back-up are poorly
based.
Turning to commercial shipping, it is here without doubt
that GPS has made the biggest inroads. No doubt the
reason has been that there is an IMO requirement for
ships to carry either a “GNSS” or, since one is not yet
available, an equivalent which since the demise of Omega
has been universally interpreted to mean GPS, although
IMO does not mention GPS by name. Practically every
ship now has GPS on board for deep-water navigation,
and not merely one; most have two and sometimes as
many as six. GPS is also used in more demanding
situations where considerable accuracy is required, such
as berthing, although it is then used in differential mode
rather than raw. The question of a back-up thus might
seem rather more important for commercial shipping than
for any other class of user, but let us examine the problem
in more detail.
Before GPS nearly every European vessel had Decca
Navigator and swore by it as the most accurate and
reliable system of its day.
Fig. 7 Decca Navigator Mk 21 Marine Receiver.
That didn’t stop them deserting it in droves once they
discovered GPS, making Decca unviable commercially
and eventually leading to its switch-off. One of the main
users of Decca were fishermen but they similarly went
over to GPS once they found they could buy a GPS for
less than a Decca. They didn’t seem to have much
concern about keeping it as a back-up for GPS even
though they had already installed and paid for it. This
lack of protest over the demise of Decca demonstrates
that there is really not much concern amongst these users
about the availability of a back-up for GPS. No doubt
they consider that if they were out at sea and lost their
GPS they would simply revert to long-standing
traditional m e tho ds o f navigation.
The loss of GPS while on the oceans would not actually
be a matter of deep concern to the average Master;
standard navigation could still be carried on quite
successfully by traditional methods and the only loss
would be perhaps a certain amount of extra fuel used. In
close waters a Masters’ main concern these days is
avoiding other shipping and for this purpose his radars
are his main tool. They also serve admirably as a means
of avoiding land and other obstacles and do a job no GPS
or GNSS could do.
More critically, some vessels, for instance high-speed
ferries, now use differential GPS for berthing and its
sudden loss could be serious. However, they still station
lookout men at each critical point on the vessel and
berthing without GPS is regularly practised. Its loss
would not stop operations although it might slow them
down. It should also be remembered that if it were
desired to continue operations uninterruptedly should
GPS disappear it would be essential that any backup had
the same precision as the local special-purpose DGPS
systems they use, claimed to be accurate to the region of
half a metre. There is no current marine radio aid that
can achieve this, differentially operated or not, so the
selection of a backup radio aid would be extremely
difficult.
Some of the most critical users of GPS are the marine and
land surveyors. Both have adopted it because of its
uniquely high accuracy, particularly in differential mode,
and its height-measuring capab ility.
There are no other systems that would enable them to
continue operations uninterruptedly; no land-based
system can measure height or get anywhere near DGPS
accuracy. For this reason its loss would be a very serious
matter for them that could not be solved by any current
standard system. However, neither operate in “loss-of-
life” situations and, although costly, could suspend
operations temporarily. The argument for a backup here
is not technical – it is economic.
Then we have the proposed use of GPS for road-tolling;
tracking offenders, and similar “Government-inspired”

Blanchard: Would a GNSS need a backup? 313
ideas. There is actually a much bigger problem here than
whether the signals might be lost - it is whether this sort
of use is legal. If a government says to its citizens “We
intend to make a law that you must fit GPS so that we can
track you and charge you for using roads/going too
fast/parking in the wrong places, etc.” then it is legitimate
for the citizen to ask what measures the government has
taken to ensure GPS will always be there providing the
necessary accuracy to do these things. Has it obtained a
guarantee from the Americans that they will never alter it
in any way without first consulting them about what
effect it might have on their legislation? Of course not,
and unless the USA changes its attitude towards foreign
users very radically it never will. Accordingly any such
use can only be on an experimental basis and that being
so there is no need at all for a backup.
Lastly we come to the timing business. Most data systems
rely on accurate timing, including much radio and
television broadcasting as well as cellphones and broad-
band web data. Their basic timing is often derived from
standards incorporating a GPS-disciplined reference
oscillator but it is incorrect to claim that were GPS to
disappear these standards would become unusable. Their
specification usually includes the ability to operate
sufficiently accurately for a week or more without GPS
correction and altho ugh this does not cover complete and
permanent loss of GPS it certainly covers short periods of
accidental or deliberate jamming. In really critical
applications timing is derived from an atomic standard
that does not require GPS correction.
We see therefore that although individual users of GPS
might want a backup they really cannot have it both ways
– GPS is a totally free service provided by the generosity
of a foreign government, and they cannot blame their
own governments for not using taxpayers money to
backup a system which they did not install or authorise,
and do not control.
Actually, many Governments have recognised the reality
of the situation; that is, that a large number of navigators
will be using GPS regardless, and have installed
Government-run differential and augmentation systems.
Thus, IALA (the International Association of Lighthouse
Authorities) has encouraged its members to use their old
MF radio-beacon sites to provide differential services,
and general-purpose systems like WAAS and EGNOS,
based on satellites, have appeared.
Fig. 8 EGNOS system (from an EC diagram) .

314 Journal of Global Positioning Systems
Fig. 8 EGNOS system (from an EC diagram) .
Fig. 9 EGNOS cover (from an EC diagram).

Blanchard: Would a GNSS need a backup? 315
These systems monitor GPS continuously and transmit
correction information so that if a GPS satellite fails it is
instantly apparent. They also transmit other information
enabling users to obtain improved accuracy. These
systems are not fall-back systems, in that if GPS fails
totally they could be used to continue navigation, but at
least they are an independent check on its performance.
4. THE GNSS SITUATION – DOES IT DIFFER
FROM GPS?
Does any of this change when we start thinking about a
GNSS instead of just GPS? Yes, it does, in quite major
ways, but perhaps not those we might expect. A GNSS
will be an integrated satnav system made up of several
components owned perhaps by different bodies but under
the control of one central authority. The one most often
talked about is a combination of GPS with Galileo, the
European system, so let’s have a look at that.
GPS, it cannot be said too often, is a military system with
civil users allowed on sufferance.
The main driver behind Galileo is civil control with the
military allowed in on sufferance. To combine the two
under a common set of rules will require the very
considerable reconciliation of US military with European
civil views which, as we have seen repeatedly quite
recently, is extremely difficult to attain and may never
happen. The likelihood is that the US will continue to
operate GPS under its own military control and Galileo
will be operated mainly under some form of international
civil control. “Civil control”, in this context, means
amongst other things that there is full disclosure of basic
operating engineering parameters and standards in order
that specifications can be drawn up for safety standards
and minimum operating performance specifications. That
has not occurred for GPS and is not likely to. It is why
ICAO has never certified it for sole use. That being so,
any civil use for safety-critical purposes of a GNSS that
combines the two could only be done on the basis of what
Galileo alone offers, thus negating the whole purpose of
integration. It might as well be done completely ignoring
GPS from the outset. This is so irrespective of whether
they are technically integrated; that is not particularly
difficult and in fact has already been agreed formally.
The proponents of Galileo have said from its inception
that it must be “inter-operable” with GPS and technical
co-operation over such things as modulation, coding and
frequencies has been excellent.
Quote:
“The Council agrees :
......................................................
That Galileo should be interoperable with existing
satellite navigation systems ; it should in particular be
interoperable with GPS and its successor systems
through an EU-US agreement that should be negotiated
as soon as possible”
(Council of European Transport Ministers Statement, 26th
March 2002)
Since then, a formal agreement has been concluded. They
can now co-exist on the same frequencies and there is no
chance that one will jam the other. It will not be difficult
to make a receiver that will receive both simultaneously.
Both will provide a free service and for those users who
do not have to use satnav for mandatory purposes it will
be marvellous to have upwards of 60 satellites in the sky
and performance in cities and other obstructed situations
will be much improved. But the problem of legal recourse
will still remain for the operating authorities who must
consider how to reconcile GNSS with their existing
regulatory structures which have been bu ilt up ov er many
years and are so embedded they are extremely difficult, if
not impossible, to change.
It has already been emphasized that there is a great
difference between non-mandatory and mandatory users
but there is no harm repeating it. We must be very careful
to draw a distinction between non-mandatory systems and
those fitted to ensure compliance with regulations of one
type or another. Once the fitting of a system is required
by law a whole set of new parameters appear in which a
prime consideration is who owns, operates and controls
the basic system. Who is it that guarantees its
performance? You cannot make a law requiring the
fitting of a system you cannot guarantee. There are no
guarantees for GPS; to be sure there is a set of basic
operating goals set out in documents like the “SPS
Performance Standards” document but they are not
guarantees. This is one of the main reasons for Galileo,
which, being fundamentally under civil control will have
the background legal structure to enable guarantees to be
given.
It may be thought too much emphasis is being laid on this
aspect. After all, a prime example of a radio navaid that
was never under direct legislative control but nevertheless
widely used was the Decca Navigator system. At its peak
one of the most widely-used marine radio navaids in the
world it never came under Government or international
regulation and was operated solely on the basis of
common commercial prudence. But because of this lack
of regulation it was never adopted as a mandatory navaid
in either aircraft or ships and therefore whether one was
carried or not was entirely at the discretion of the Master.
Over 30,000 vessels and 10,000 aircraft decided it was
worth the risk and fitted it. Decca was so successful
because it filled a gap in the armoury of marine navaids
and was operated by a commercial company that had to
be responsive to its customer’s wishes. There were free
alternatives to Decca in the form of Loran and Transit but
although they were sometimes fitted the same ships

316 Journal of Global Positioning Systems
almost always had Decca as well. Decca became a
byword for reliability and performance and those are
attributes valued by a navigator above almost all else.
The key was that it was a commercially responsive
company that ran it and spent a great deal of effort
ensuring its customers were happy with it. Not so with
GPS. It is not a commercial offering and the aims of its
owners ar e not primarily to satisfy civil users. Consider a
little recent history. Most years, there are during the
Spring months significant periods – half-an-hour or more
- on some days when GPS satellite availability is quite
marginal over the UK - only four satellites at elevations
above 20° and in poor positions for good fix accuracy.
Fig. 10 GPS availability, UK.
This might seem a high cutoff elevation but it is one
many people are using and is reasonable for cars
travelling in built-up areas where even higher cut-offs
might be necessary - 40° has been quoted. These four
satellites often on ly provide an HDOP of 20 or so (35 has
been seen on occasion) and the result is very poor
accuracy - 20 0 m et r e s or more.
Fig. 11 Poor GDOP

Blanchard: Would a GNSS need a backup? 317
On these occasions there was nothing wrong with the
GPS system; there were no satellites off-air; no
maintenance was being done; and there was nothing in
NANU’s about it. These periods of poor HDOP were
within the published performance limits for the GPS
Standard Positioning Service and therefore there was no
cause for complaint. The problem was just a fortuitous
combination of satellite positions caused by some
satellites not being quite on station and others being
drifted around their orbits to new positions. Only
Northern Europe was affected and then only for periods
less than the outage periods permitted by the statistical
availability parameters. It is entirely possible that this
could recur in future - and there is absolutely nothing to
be done about it! Would an EU Government be prepared
to have its expensively-implemented road pricing scheme
collapse even for half-an-hour at unpred ictable times?
While this is true of GPS, one of the major drivers, if not
the main driver, behind Galileo was to remedy this lack
of a GPS performance guarantee by initiating a civil-
based system operated with civil users as its main
concern and totally resp onsible to them.
Fig. 12 The Four Galileo Drivers.
If it were to be used for road-pricing then no doubt
extreme care would be taken to ensure that this sort of
situation did not occur and legal guarantees would be
much easier to enter into. Although the military will also
use Galileo, (under the guise of a “Government Service”),
they will not be in the same powerful position as in GPS
since they will not be providing the major funding,
probably much to their relief! The PPP concept is
designed in large part to provide some degree of
Governmental control over a privately-funded enterprise
and is being actively discu ssed for Galileo. That is not to
say there are no problems at all using Galileo but only
that they will be present to a lesser degree.
5. COULD A L EG A L LY-ENFORCEABLE
GUARANTEE REGARDING GPS EVER BE
GIVEN?
The US DOD, like any other national military
organisation, is charged with the defence of its national
citizens, not the provision of civil services even to its
own citizens. That is what other Government departments
are for. So, although there might be an understanding
between civil and military within the US itself, possible
because they both operate under the same fundamental
legal system, there is no chance that that understanding
could be extended to foreigners who considering it at the
most venal level contribute nothing in financial terms.
The DOD is not a commercial organisation committed to
providing value for money, and foreign part-control of

318 Journal of Global Positioning Systems
one of its vital systems would, rightly, be anathema to it.
There have been attempts in the USA to persuade us that
GPS is not really under military control and is fully
responsive to civil needs. It is pointed out that it is
managed by IGEB, an organisation on which 7 US civil
GPS-user agencies are represented as against only 2
military, and is operated by the US Air Force only on
behalf of the Government as a whole. This is very
commendable but one might ask where is the foreign civil
representation? Of cour se, there isn’t any, an d one would
not expect there to be - it is a US National system, not an
international one. And that is the problem.
Fig. 13 GPS control
Sovereignty is jealously guarded by all countries. After
all, without it a country is not in co ntrol of its own affairs
and subordination of sovereignty to a foreign power,
which is what legislation for the use of a foreign-owned
system amounts to, has no historical precedent in any
major country. For instance, the UK Civil Aviation Act
of 1921 was passed with the express object of ensuring
the UK maintained complete so vereignty over its airsp ace
as well as land, a principle maintained rigorously by
every country ever since. We need only look at the
problems still being encountered with attempts to
establish a common air traffic control system for Europe
to see the complications even a slight derogation of this
principle can cause.
6. WILL GALILEO ALONE FILL THE BILL?
We are therefore in a situation where it is impossible to
certify GPS for use in any regulated civil situation and if
this situation persists we will have to rely totally on
Galileo. That means that Galileo could NOT be
reinforced by GPS and we would still have only the 30
Galileo satellites in the sky capable of being used for
these purposes. What then of the much-vaunted
reinforcement of GPS by Galileo in city canyon areas on
which so much depends? We would simply have the
same situation as today, not enough satellites.
Galileo at least starts off with the premise it is a civil
system setting out to satisfy civil needs.
That being so, all the usual panoply of civil and
commercial procedures and safeguards can apply. If a
proper Public Private Partnership is forged, in which a
not-for-p r ofi t , pr o babl y quasi-Governmental, orga ni sat i o n
oversees the safety and regulatory aspects while private
enterprises seek profit in the system where they may, it
should be possible to ensure service guarantees sufficient
to enable legislation for its use to be feasible. The
difficulty Galileo will have is in the formation and
working of the overseeing authority. Since aviation will
be only a minority partner it cannot be the sole or even
the major determinant of policy and an organisation such
as Eurocontrol will be inappropriate. However, there is
plenty of precedent for a new organisation to be formed
along the lines of the European meteorological and
telecommunications satellite consortia EUMETSAT and
EUTELSAT; EUNAVSAT perhaps?

Blanchard: Would a GNSS need a backup? 319
Fig. 14 Galileo aims.
The EU need not prevent non- EU countries joining in, as
already demonstrated by non-EU participation in these
and other organisations. Although no doubt the usual
tedious and lengthy consultation would be involved there
appears to be no really fundamental problem preventing
such an organisation being formed. Some political and
sovereignty issues would still remain, of course, but in a
less severe form considering the existence of the
European Union. The big difference between these older
organisations and EUNAVSAT would be that are no
legal compulsion s surrounding them. It is not yet a jailing
matter if you do not use EUTELSAT or EUMETSAT,
but it might well become one if you refuse to have
EUNAVSAT in your car!
But a word of warning. Little is being said publicly ab out
what type of organisation might eventually run Galileo,
for all sorts of very good reasons. One is that it will take a
very long time to thrash out and it would be premature to
go on the record now. So the likelihood is that the initial
operation of Galileo will be under the au spices of ESA as
a test and trial system but for how long will that last? 20
years? While it is in the status of a T and T system no
legislation at all could be passed so it might not be until
2030 or so we will have a fully civil-responsible Galileo
run by a legally-responsible organisation and all its
putative advantages secured.
7. 2010, GALILEO, GPS III, AND ALL THAT.
Let us ignore such gloomy forebodings and assume that
in 2010 we have a civilly-responsible Galileo system
around which legislatio n could be framed but with which
GPS is unusable in any leg islated situation . Unfortunately
it will be a Catch-22 situation because the technical
problems we already have with GPS will still exist - a
restricted number of satellites causing non-performance
in city areas - so no performance guarantee to adequate
standards will be possible and that would prevent
legislation anyway. And what of the millions of drivers
who would have GPS-only receivers - are they to be told
to get rid of them and fit new, certified, Galileo-only
models when they would not see any significant
difference in performance? At, possibly, A$2000 a time
(it would be claimed certification costs are not negligible)
when they can get GPS sets for a few hundreds?
Aircraft do not experience the same restrictions on
satellite visibility as cars and provided the legal situation
is resolved the use of Galileo for approach and landing
could, in principle, be authorised. So it might - for
European and other Galileo-subscribing non-US
commercial air traffic, but what about intercontinental
traffic? The US would no doubt not want to authorise
Galileo, not having any control over it over their own

320 Journal of Global Positioning Systems
territory, in the same way Europeans could not authorise
GPS. But if Europe insisted on legislating for Galileo the
result would be that international traffic would have to
carry dual capability and separate GPS/Galileo
equipment instead of a combined GNSS box. This would
come about because mixing a certified with a non-
certified system is strictly a no-no. This might not matter
as far as cockpit operation was concerned but the
important point is that they would be carried not to
reinforce one another but so that the appropriate system
could be switched in when over different countries. If an
aircraft had an incident in the USA while using a dual
GPS/Galileo equipment that integrated the two instead of
treating them separately the US authorities might wash
their hands of the problem on the grounds that a non US-
approved system was in use, and the reverse might
happen in Europe. This is the exact situation airlines now
fear and is why there will almost inevitably be a strong
reaction if it is ever proposed that GPS/Galileo or any
other satnav system should be legalised as a sole system.
In the face of this, satnav would stay where it is now; a
useful support system fo r INS but not much else. Since it
would itself then be in the position of a back-up for other
aids there would not be much point installing a back-up
to back-up the back-up.
8. A BACK-UP FOR A GNSS?
If the above is correct then the obvious inference is that
we will never have a true GNSS. There will be GPS,
Galileo, perhaps Glonass, possibly all integrated together
at the engineering level and receivable on one single
receiver, marvellous for the non-mandatory user but
useless when legislation is involved. It may well
eventually be possible to legislate around Galileo, since
that is one of its prime purposes, but hard ly around either
GPS or Glonass while they maintain their military stan ce.
So the question is moot, there will not be a GNSS to
back-up.
Fig. 15 Requirements f o r a GNSS.
9. A BACK-UP FOR GPS OR GALILEO, THEN?
We have seen above that a back-up is required only for
those for whom satnav is an essential tool, for the others
it is only a matter of convenience. But it turns out that
neither of the major professional users, aviators and
mariners, require a back-up because they already have
well-proven systems in place; VOR/DME/ILS for one

Kim et al: A Step, Stride and Heading Determination for the Pedestrian Navigation System 321
and radar for the other. For “voluntary” users, no back-
up is required either because it is not generally a safety
matter. The cases of amateur sailors and aviators who
might get lost if they suddenly lose their satnav are the
“hard cases” that make bad law. What is it that might
cause them to lose their satnav anyway? Ignoring
sillinesses like failing to carry spare batteries there is
actually very little. The hard case we think about is where
satnav has been working perfectly well on a trip and it
suddenly disappears because all the satellites have been
switched off. This is actually an impossibility because
they could not be made to all disappear instantly; they
would disappear one by one as they came within range of
command stations that could command a switch-off. And
it would be done only in a “grave emergency”, we are
told, requiring an international situation so bad that it
would have been general public knowledge well in
advance and the possibility of a GPS switch-off well-
known. As regards a total failure of the GPS command
system, that is about as likely as all four engines failing
simultaneously on a 747. Everything is duplicated, if not
triplicated, and even if the command centre in Colorado
were to fail completely, there is a reserve one in
California. Failures in the satellites themselves are
certainly possible and have occurred. The guard against
these is proper monitoring and Galileo will have much
improved built-in monitoring compared with GPS, the
key being instant no tification to users of a failure through
the satellite signals themselves transmitted from other
satellites.
What about “natural” causes? There are no significant
propagation effects that could affect all visible satellites
simultaneously. Scintillation can occasionally cause
temporary fades of individual satellites but it is a very
localised effect and there are always sufficient satellites
in other parts of the sky. There are no precipitation static
effects or skywaves as with low-frequency systems. The
only significant impediments to propagation are blockage
by buildings, trees, and other physical objects and the
remedy is obvious.
The ease of jamming the relatively weak signals emitted
by satellites is often quoted as if no other radio signals are
that weak. To take what might be called “official”
jamming first, it is well-known that the preferred
“official” method of denial of GPS service to an enemy is
by local jamming to wipe out the SPS while leaving the
PPS unaffected. If things are bad enough to require this
then perhaps the non-military user should not be in the
area in the first place, but in any case the possibility will
have been well-publicised. Secondly, it has been claimed
that a terrorist, or, perhaps worse, an electronic hobbyist
would find it easy to jam GPS for his own purposes. The
first thing to be said about this is that it would not affect
professional users in the slightest - no aircraft or ship
would be totally dependent on satnav as we have seen.
The amateur seamen and aviators who might be relying
on it rather more would not necessarily be affected very
greatly either, since unless it were a major and very
professional attack, they would eith er be out of range of a
land-based jammer or would fly through the area of
jamming relatively quickly. It is only local land-based
users who would be really affected and for them there is
no safety issue, on ly incon v enience. “Ho bbyist” j a mmers
only do it for some easily-visible effect, but unlike
jamming communications, jamming a navigation system
does not show any su ch effects. It would be quite easy to
jam an ILS for instance, with possibly more visible and
instant effects even if it is only the sight of a jumbo
having to go round again, but there have not been any
recorded cases so far. As regards a radio-based backup
for satnav to guard against jamming, that too could very
likely be easily jammed. Loran- C has been quoted in this
context but in fact over much of its coverage area its
signals are not at all strong and they are just as liable to
malicious jamming.
10. CONCL U SIO N S.
The possibility of total and instantaneous failure of a
satnav system like GPS or Galileo is so remote as to be
discounted. No other radio navigation system has ever
had the advantage of multiply-redundant independent
transmitters unaffected by power supply outages, storm
damage and insurrection. Natural causes of loss of signal
are rare and are guarded against by the multiple satellite
sources. Such possibilities of signal loss as there are
come mainly from human activities that are limited in
their effect. The reliability that results from these factors
is such that no ground-based radio aid could possibly
approach it and it would be ridiculous to have a back-up
radio system that was less reliable than the system it was
supporting.
On examination, it is seen that in fact there are really
very few users who actually need a new type of backup
for satnav anyway, whether the satnav is GPS, Galileo or
some combination that might be called a GNSS. For
those for whom the loss of satnav would be important
there are alternatives already available; for many others it
is not a safety-of-life matter. It is hardly an effective use
of resources to install a completely different system that
would almost never be used. The important matter is not
a back-up but putting into place the correct international
arrangements to enable the genuine co-use of GPS,
Galileo, and Glonass as one truly integrated GNSS
system.