An Introduction to Direction Finding

Methodologies
White Paper

Please find the most up-to-date document on our Direction Finding Resources page:
https://www.rohde-schwarz.com/direction-finding
Paul Denisowski
1.2020 – 1.00
White Paper
 Table of Contents

Table of Contents
1 Basic concepts ........................................................................................ 4
1.1 What is Direction Finding? ............................................................................................... 4
1.2 Power of Arrival vs. Angle of Arrival ............................................................................... 4
1.3 Determining bearings ........................................................................................................ 5
1.4 Using bearings ................................................................................................................... 6
1.5 Manual vs. Automatic Bearings ....................................................................................... 6
1.6 Multipath ............................................................................................................................. 7

2 Manual direction finding ......................................................................... 8

2.1 About manual direction finding ........................................................................................ 8
2.2 Antennas used in manual direction finding .................................................................... 8
2.3 Practical considerations in manual direction finding .................................................... 9

3 Doppler ....................................................................................................10
3.1 About Doppler direction finding ..................................................................................... 10
3.2 Creating a Doppler shift .................................................................................................. 10
3.3 Implementing Doppler antennas .................................................................................... 11
3.4 Practical Considerations in Doppler DF ........................................................................ 12

4 Watson-Watt ............................................................................................13
4.1 About Doppler direction finding ..................................................................................... 13
4.2 About Adcock Antennas ................................................................................................. 13
4.3 Implementation of Adcock Antennas ............................................................................ 14
4.4 Practical Considerations in Watson-Watt DF ............................................................... 14

5 Correlative Interferometry ......................................................................16

5.1 About Correlative Interferometry ................................................................................... 16
5.2 Determining bearings using correlative interferometry .............................................. 16
5.3 Implementation of CI antennas ...................................................................................... 18
5.4 Practical considerations in correlative interferometry ................................................ 18

6 Time Difference of Arrival ......................................................................19

6.1 About time difference of arrival ...................................................................................... 19

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 2

 Table of Contents

6.2 Using hyperbolae ............................................................................................................. 19

6.3 Implementation of TDOA ................................................................................................. 20
6.4 Correlating signals .......................................................................................................... 20
6.5 TDOA sensors .................................................................................................................. 21
6.6 Location coverage and accuracy ................................................................................... 22
6.7 Practical considerations in TDOA .................................................................................. 22

7 Hybrid Methodologies ............................................................................23

7.1 What is a hybrid methodology? ..................................................................................... 23
7.2 Combining angle of arrival and time difference of arrival ........................................... 23
7.3 Advantages of hybrid methodologies ........................................................................... 24

8 Conclusion ..............................................................................................25

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 3

 Basic concepts

1 Basic concepts
1.1 What is Direction Finding?
In the radio-frequency world, the term "direction finding" (DF) refers to the use of
specialized instruments, antennas, and methodologies to determine the physical or
geographical location of an emitter; that is, a source of radio-frequency energy. Although
these methodologies differ substantially in terms of how they estimate an emitter's location,
they all share the common goal of the highest possible accuracy. Accuracy requirements
do however vary between applications. For example, finding a ship in distress on the open
water only requires an accuracy of several hundred meters, particularly in good weather.
On the other hand, eliminating sources of radio-frequency interference often requires
accuracy of several meters or less. Note too that in some cases, direction finding targets
are “non-cooperative.” In other words, not only do they not want to be found, but they may
also take steps to hide their location or otherwise complicate the direction finding process.
This whitepaper discusses the basic principles behind radio direction finding and the most
commonly used direction finding methodologies: manual angle of arrival (AOA), Doppler,
Watson-Watt, correlative interferometry (CI), time difference of arrival (TDOA), and hybrid
methodologies. A general technical introduction is provided for each of these
methodologies, including an overview of how they are implemented and the relative
strengths and weaknesses of each.

1.2 Power of Arrival vs. Angle of Arrival

Before discussing individual methodologies, it would be helpful to define some terms that
are often a source of confusion when discussing direction finding: namely, the difference
between “power of arrival” and “angle of arrival." There are various competing definitions
for these terms, but in this whitepaper, "power of arrival" will be used to describe
measuring the level of received
radio-frequency energy at a
given location. In this sense,
power of arrival could be used
to locate an emitter by simply
moving our receiver and
measuring the received RF
level to create a "heat map"
(Figure 1) of power versus
location. The points at which
the highest power levels were
measured are the closest to the
emitter location. This Figure 1 - Heat map (power of arrival)

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 4

 Basic concepts

procedure, strictly speaking, is

not direction finding since we
are not finding the direction
towards the source, but rather
a more general case of
radiolocation.
"Angle of arrival," on the other
hand, determines the angle,
that is, the bearing (or
azimuth) towards the source of
RF energy. This is true Figure 2 - Bearing (angle of arrival)
direction finding in the sense
that we determine the direction from which the signal is arriving (Figure 2).
This distinction is important because the vast majority of direction finding methodologies
are based on angle of arrival, not power of arrival. Angle of arrival allows the use of a wide
variety of methodologies, all of which are faster, more efficient, and much more accurate
than approaches based on power of arrival.

1.3 Determining bearings

Angle of arrival based direction finding methodologies all share one common task:
determining the direction (angle, bearing) from which a signal is arriving. Recall that there
are three things that can potentially change as a radio-frequency signal propagates
through space: its amplitude, its frequency, and/or its phase. Since these changes are
primarily a function of the path between transmitter and receiver, direction finding
methodologies can calculate bearings using these location-dependent variations in the
received signal. Direction finding methodologies based on angle of arrival use one of these
variations to compute bearings: changes in amplitude, frequency, or phase.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 5

 Basic concepts

1.4 Using bearings

These bearings, which
represent the direction
towards the emitter, can be
used in two main ways. The
first of these involves a single
bearing. A single bearing
provides direction but not
distance information, so the
main application of a single
bearing is homing towards a
target – we simply follow the
bearing until it leads us to the
source. The second case
involves multiple bearings. If
we are able to take bearings
from multiple locations, these
can be plotted together to Figure 3 - Triangulation
compute the most probable
location of the target – a process often called triangulation (Figure 3). Triangulation is a
simple mathematical calculation whose accuracy is entirely a function of the accuracy of
the bearings used to produce it: accurate bearings yield accurate triangulation results.
In most cases, three or four good bearings are sufficient for obtaining a good triangulation
result. Further increasing the number of bearings used in the triangulation calculation
above more than five or six does not typically yield additional accuracy. It is also worth
noting that “real-word” bearing results taken from multiple locations will rarely all intersect
precisely at the same point. In some cases, it is also necessary to discard obvious fliers or
so-called “junk bearings” that clearly are not pointing towards the true source of the signal.
These can be caused by multipath or reflections of the RF signal, which will be discussed
later.

1.5 Manual vs. Automatic Bearings

The methods used to obtain bearings can be divided into two general categories. The first
method involves some type of manual process. In almost all case, this manual
methodology means that the operator manually moves or points the antenna until the
strongest signal level is observed. Note the use of the word “observed” – in most cases, a
human operator that makes the determination as to when the antenna is pointed in the
“right” direction. This means that manually obtained bearings tend to be somewhat
“operator-sensitive,” which can be either good or bad, depending on the operator’s skill
level and experience. In the proper environment and when performed by a skilled operator,
manually obtained bearings can be highly accurate. The other method of obtaining

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 6

 Basic concepts

bearings involves the use of automatic direction finding methodologies. Here, a system
*automatically* determines bearings based on one or more changes in the signal
characteristics (amplitude, frequency or phase) at a given location. One of the main
advantages of automatic direction-finding methodologies is that they do not depend as
strongly on the operator's skill level. Another important advantage is that these automated
methodologies can produce accurate results even in challenging propagation
environments, particularly in the presence of multipath.

1.6 Multipath
As the name implies multipath
means receiving a signal from
multiple directions
simultaneously (Figure 4).
Radio frequency signals can be
reflected from radio-opaque
objects, such as concrete,
metal, etc. and this means that
the signal may appear to be
coming from the reflection
point, rather than from the
source itself. This is a
particularly significant issue in
urban or mountainous areas.
Most people have experienced
multipath in their everyday
lives: when sitting at a stop
Figure 4 - Multipath light or stop sign, an FM radio
station signal may get stronger
or weaker as the car is inched forwards. The change in the signal strength and quality are
due to the multipath profile at that particular point - moving the car a few inches can
dramatically change the nature of the received signal. In direction finding, having a signal
that appears to come from multiple locations or which changes level due to constructive or
destructive interference is a significant problem, and therefore multipath is usually the
single biggest challenge in direction finding. In many cases, a “good” direction finding
system or methodology is primarily dependent on its ability to obtain good bearing results
even in the presence of substantial multipath.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 7

 Manual direction finding

2 Manual direction finding

2.1 About manual direction finding
Manual direction finding is the most basic direction finding methodology. Recall that all
direction finding methodologies based on angle of arrival use a change in amplitude,
frequency, or phase to determine the direction towards a signal. Manual determination of
angle of arrival is based on amplitude comparison. In manual direction finding, a
directional antenna is physically moved or rotated until the system operator obtains the
maximum received signal strength, usually via visual inspection. This direction, or bearing,
is then used either for homing or for triangulation. The directional antenna can be either
handheld or mounted on a tripod, mast, or vehicle.

2.2 Antennas used in manual direction finding

Although a directional antenna
is a requirement for direction
finding using manual angle of
arrival, the type of directional
antenna will vary based on the
application. Some antennas,
such as Yagi antennas, may
have very good directionality
(narrow beamwidth), but also
numerous sidelobes and a
limited usable frequency range.
Log-periodic antennas (Figure
5) can cover wide frequency
Figure 5 - Log-periodic antenna used in manual DF ranges with minimal or no
sidelobes, but the main lobe is
often quite wide. Horn antennas are excellent (and sometimes the only choice) at very
high frequencies, but can be challenging to work with. Interaction between the antenna
and the mounting hardware, including the arm and body of a human operator, can also
affect results in a non-trivial manner.
In most cases, the angle or bearing is normally determined and/or recorded using a
manual or electronic compass, relative to magnetic north, since this facilitates the use of
these bearings in mapping and triangulation.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 8

 Manual direction finding

2.3 Practical considerations in manual direction finding

The biggest advantage of manual angle of arrival is that it is low-cost and portable,
especially when done using handheld antennas and portable receivers. Even when using a
tripod or other mounting device (Figure 6), manual systems are almost always man-
portable and can be quickly deployed in almost any location with minimal setup time.
Vehicle-mounted systems can range from the simple "Yagi-on-the-roof" to more elaborate
magnetic mounts and elevating masts.
Manual angle of arrival-based
approaches can often yield acceptable
results, depending on the type and
nature of the target signal and the
ambient propagation environment.
That said, there are a number of
limitations when it comes to manual
direction finding approach. As
mentioned earlier, the effectiveness of
this methodology can be strongly
dependent on the operator’s skill level
and experience since the antenna is
usually being manually manipulated
and the bearing determination is being Figure 6 - Tripod mounted manual DF antenna
manually made by a human operator,.
Accuracy can be poor for distant targets due to the human aspect as well – even with a
tripod and a very narrow antenna pattern, the amount of human-introduced bearing error
can lead to substantial location errors at distances of more than a few hundred meters.
Antenna choice can be problematic, especially when it comes to balancing narrow
bandwidth and narrow beamwidth. Lastly, manual angle of arrival techniques do not work
well when trying to locate short-duration signals. A signal that only appears for a few
seconds at a time can be “there and gone” before the operator has time to rotate the
antenna and determine and/or record a bearing. These limitations are one of the reasons
for the development of automatic direction finding methodologies.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 9

 Doppler

3 Doppler
3.1 About Doppler direction finding
The first automatic direct finding methodology we will discuss is called "Doppler" direction
finding because it is based on the Doppler effect. The Doppler effect, or Doppler shift, is
named after Christian Doppler, who first described it in 1842. Doppler shift is a type of
frequency modulation caused by relative motion. If an object is moving towards an
observer, the frequency of waves emitted by an object will appear to increase. Similarly, if
the object is moving away from the observer, a downward shift in frequency occurs.
Doppler shift applies to many different domains. Most people are familiar with Doppler
shift in the audio frequency domain, such as the change in pitch in the whistle of a passing
train. Doppler shift also occurs in the visible energy domain, such as the red shift of stars
moving away from Earth. In the radio frequency domain, Doppler shift can be used for the
purposes of direction finding.
As we move towards a signal source, the received frequency of the signal will shift
upwards. If we move away from a signal source, the received frequency will move
downwards. If we are able to detect and measure this shift, we could then determine
whether we are moving towards or away from the signal source. This means we could get
a direction. Thus, in order to use Doppler shift for direction finding, the receiver, or more
precisely, the antenna, needs to be moved in such a way that a measurable Doppler shift
is created.
There are, however, a couple of challenges with regard to this approach. First, if the target
were stationary, the receiver would have to be moved to create a Doppler shift. There is
also the question as to whether it is possible or practical to move the receiver fast enough
to create a measurable Doppler shift.

3.2 Creating a Doppler shift

Fortunately, it is not the receiver but rather the antenna that needs to be moved, so the
problem now becomes: how do we move an antenna in such a way as to create Doppler
shift? We will begin with a single antenna mounted on a rotating disk (Figure 7).

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 10

 Doppler

Figure 7 - Rotating disk concept

As this disk is rotated, the antenna will move closer to

and then further away from the transmitter. At
positions A and C (Figure 8), the antenna is stationary
relative to the transmitter – that is, the antenna is
neither moving towards nor moving away from the
transmitter. This means that the Doppler shift is zero
at positions A and C. At position B, we have the
maximum speed moving away from the transmitter
and at position D, we have maximum speed moving
towards the transmitter. The varying Doppler shift
creates a so-called Doppler sine wave, which is a
maximum when the antenna is at position D, with
zero crossings at positions C and A. The second zero
Figure 8 - Doppler Sine Wave
crossing at point A represents the position or angle
that points towards the transmitter.

3.3 Implementing Doppler antennas

The problem with this implementation as described is that an antenna on a rotating disk is
not practical. The rotational speed required to create a measurable Doppler shift is much
too high to be implemented as a vertical antenna on a physically spinning disk. However,
we can simulate a rotating disk by using an array of antennas, usually four, and then
rapidly switching between them. Each of the antennas is used to generate a series of
Doppler pulses, and the system uses this information to synthesize the Doppler sine wave.
Although the switching between the antennas must be very fast in order to accurately
synthesize the Doppler sine wave using inputs from four discrete antennas, this is not a
technological challenge.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 11

 Doppler

As mentioned above, most Doppler DF

antennas have four elements, equally
spaced, with vertical polarization (Figure 9).
In many cases, Doppler antennas are
vehicle-mounted, but Doppler antennas can
also be fixed-mounted or configured in a
man-portable package. It is possible to have
Doppler DF antenna arrays with much more
than four antenna elements. The
hypothesis here is that the synthesized
Doppler sine wave will be more accurate if Figure 9 - Doppler antenna with four elements

more measurement points are used. While it

is in fact true that a larger number of antennas improves Doppler DF results, this only
applies if we also increase the diameter of the antenna array. It is also worth noting that
some Doppler systems have multiple sets of antenna elements to cover wider frequency
ranges. For example, a Doppler DF array with eight elements may simply be two arrays of
four elements each – a set of four longer antennas for VHF and four shorter antennas for
UHF.

3.4 Practical Considerations in Doppler DF

Doppler DF is relatively low cost compared to other automatic DF methodologies. Part of
the reason why Doppler is so low cost is that it’s possible to build a Doppler system using
simple vertical antennas and an off-the-shelf receiver, although commercial Doppler DF
systems usually have a specialized DF receiver for better performance. There are
however a few limitations inherent to Doppler DF. First, Doppler DF systems more or less
require that the target signal be a constant, CW or similar type signal. Doppler does not
work well for locating intermittent, that is, on and off signals, or very broadband or noise-
like signals. The antennas or antenna elements in most Doppler systems also limit the
useful frequency range for Doppler DF. Almost all Doppler systems are designed to work
at VHF or (low) UHF frequencies, so Doppler is not a good choice for direction finding at
either HF or microwave frequencies. In addition, because Doppler antenna arrays are
made up of vertical elements, these arrays do not work well when trying to locate
horizontally polarized signals – the loss from cross-polarization limits can easily be 20 to
30 dB. A Doppler-based automatic direction finding systems is something of an “entry
level” DF system that can be a good choice for certain applications, particularly when cost
is a consideration.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 12

 Watson-Watt

4 Watson-Watt
4.1 About Doppler direction finding
Doppler can be classified as a frequency-based direction finding methodology: changes in
frequency are used to determine the direction or bearing towards the emitter. Watson-
Watt, on the other hand, is an amplitude-based DF system.
More precisely, Watson-Watt is an amplitude comparison system. It’s one of the older DF
methodologies, having been developed shortly after the First World War, and is named
after its inventor, Sir Robert Alexander Watson-Watt, one of the pioneers in the field of
radar. The Watson-Watt direction-finding methodology is based on the properties of a
special type of antenna, commonly referred to as an Adcock antenna.

4.2 About Adcock Antennas

The classic Adcock antenna consists of four
equally spaced vertical elements arranged in
pairs. This creates an antenna pattern
consisting of two figure-eight shaped lobes.
Each of the two figure-eight shaped patterns
has maximum sensitivity along the axis running
through both antennas and nulls perpendicular
to this axis. This arrangement yields a unique
set of magnitudes for every incoming signal
direction (Figure 10).
An example might be helpful: we will define two
axes, the North-South axis, for our green
antennas and the East-West axis, for our blue
antennas. Note that the nulls in these two Figure 10 - Basic Adcock antenna pattern
patterns are orthogonal, or at right angles to
each other. A signal originating due North, or with an azimuth of zero degrees, is sensed
very strongly on the North-South axis, but almost not at all on the East-West axis (Figure
11). A signal that arrives from the northeast would, ideally, result in equal measured
amplitudes for both antenna pairs. In a similar way, any arbitrary direction can be sensed
by comparing the amplitudes on the two axis. There is, however, the possibility of a 180
degree ambiguity: a signal arriving from the west and one arriving from the east will create

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 13

 Watson-Watt

the same set of magnitudes on the East-

West axis. This ambiguity is easily dealt
with using a simple sense antenna that is
usually placed at the center of the array.

4.3 Implementation of Adcock

Antennas
There are quite a few ways to physically
implement Adcock antennas. Most typically,
the individual antenna elements are
implemented as monopoles when there is a
ground plane to work against, or as dipoles Figure 11 - Bearings from an Adcock antenna
in pole-based or tower-mounted
applications. The same type of pattern and behavior can be obtained using a pair of
crossed loops as well. With regards to the arrangement of the elements, the biggest
question is how far apart to space the elements. This spacing is a compromise between
accuracy and sensitivity – placing the elements closer together gives higher accuracy, but
sensitivity is better when the elements are spaced further apart. Note that is some cases,
additional pairs of Adcock antennas can be used to overcome these constraints, but this is
uncommon in practice.

Figure 12 - Adcock antenna implementations

4.4 Practical Considerations in Watson-Watt DF

Like all other direction finding methodologies, Watson-Watt has its own particular set of
advantages, disadvantages, and practical considerations. One of the ways in which
Watson-Watt distinguishes itself from other direction finding methodologies is that it is well

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 14

 Watson-Watt

suited to HF direction finding, mostly due to the ease with which smaller antennas can be
implemented at these frequencies. Compare this to Doppler, which usually is limited to
only VHF and UHF. Watson-Watt is also very fast. There is very little “calculation” of
results required before obtaining a bearing, since the system simply compares amplitude
on two different antenna pairs. Watson-Watt has good accuracy and sensitivity, with the
accuracy depending primarily on the circularity of the antenna patterns – something that is
not difficult to control. The main drawback to Watson-Watt is that it cannot measure
elevation in addition to azimuth, and if the target is significantly higher or lower than the
plane of the antenna, accuracy can suffer.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 15

 Correlative Interferometry

5 Correlative Interferometry
5.1 About Correlative Interferometry
Manual angle of arrival and Watson-Watt are both amplitude-based direction finding
methodologies, and Doppler DF is based on changes in the frequency of the received
signal. Correlative interferometry (CI) is different from all of the previously discussed
methodologies in that it is based on changes in the phase of the received signal.
Like many other direction finding methodologies, correlative interferometry was derived
from another application, in this case, radio astronomy. Correlative interferometry
determines bearings by calculating differences in the received signals’ phase as seen at
multiple collocated antenna elements. CI antennas generally use an odd number of
antenna elements arranged in a circular pattern.

5.2 Determining bearings using correlative interferometry

In order to explain the principles behind correlative interferometry, we will use a five-
element antenna as an example. Five antennas are placed in a circular pattern, with one of
the antennas, here antenna 1, being designated as the reference channel. When a signal
arrives at the array on a given angle, for example, 45 degrees, the measured signal will be
phase offset at each of the five antenna elements. A calibration process is performed for
each reference angle, 0 to 360 degrees, in increments of one degree. The result is a table
that contains phase offsets for all incoming signal angles. Measuring phase differences is
the “interferometry” part of the methodology’s name (Figure 13).

Figure 13 - Measuring phase offsets with an interferometer

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 16

 Correlative Interferometry

When a target signal arrives at the antenna array, the received phase offsets are
measured at each antenna. The correlation between the measured phase offsets and the
calibrated or ideal phase offsets is calculated for each arrival angle. This process should
yield a clear correlation peak from which the bearing or arrival angle can be derived
(Figure 14).

Figure 14 - Deriving bearings from correlation with reference phase values

In addition to delivering a bearing result, correlative interferometry also provides a bearing

quality. Both of these correlation graphs in Figure 15 yield a peak that corresponds to a
bearing of 127 degrees. However, the graph on the left shows much stronger correlation at
this angle than the graph on the right. Therefore, while both sets of results produce the
same bearing, we would have a higher degree of confidence in one of the results than the
other one. This type of “bearing quality” result is unique to correlative interferometry.

Figure 14 - Quality of correlation

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 17

 Correlative Interferometry

5.3 Implementation of CI antennas

Antennas designed for use in correlative
interferometry usually have an odd number of
elements, typically 5 to 9, and these are
enclosed within a radome (Figure 16).
Depending on the characteristics and
placement of the antennas, a carefully
designed CI array can cover very wide
frequency ranges, up to and above 1 GHz.
CI antennas can also be implemented Figure 16 - Nine element CI antenna (radome removed)
in very small packages with good
accuracy, but the array's immunity to things such as
reflections and multipath increases as the diameter of
the antenna increases. Reflections and multipath
cause variations in the phase of the received signal.
Increasing the aperture (diameter divided by
wavelength) of a CI antenna to a value greater than
one improves accuracy since the effect of these "bent"
isophase lines is decreased (Figure 17).
However, compared to the other the DF
methodologies discussed so far, correlative
interferometry places strict requirements on the
antenna in terms of design and manufacturing
tolerance. One can “home-brew” antennas for manual
DF, Doppler, and Watson-Watt and get reasonably
acceptable results, but CI antennas require very
specialized engineering and production Figure 17 - Effect of bent isophase lines on
resources due to highly demanding design and smaller versus larger apertures
manufacturing tolerances

5.4 Practical considerations in correlative interferometry

Despite the added complexity and cost required to produce a CI antenna, correlative
interferometry has several significant advantages over other DF methodologies. The first is
very high accuracy – usually one degree or less in an ideal environment. No other direction
finding methodology can produce bearings with a similar level of accuracy. Because CI
compares phase differences on multiple elements simultaneously, it offers very high
immunity to multipath compared to other DF methodologies. Signal polarization does not
affect the accuracy of CI-based systems – this was one of the main drawbacks of Doppler
DF – and CI-based systems do not suffer from the elevation issues found in Watson-Watt
direction finding.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 18

 Time Difference of Arrival

6 Time Difference of Arrival

6.1 About time difference of arrival
Time difference of arrival (TDOA) is neither a power of arrival nor a time of arrival direction
finding methodology, but rather a time-based methodology that determines the location of
an emitter based on the time when a signal is received at multiple locations.
The basic principle of time difference of arrival is as follows: three or more receivers are
placed in different locations, and all of these receivers receive a signal. In most cases, the
distances between the transmitter and each receiver are different, and so the time at which
the signal arrives at each receiver will also be different. We can represent these time
differences as curves, or hyperbolae, and the location of the transmitter will be at the
intersection of these hyperbolae. In fact, time difference of arrival is sometimes also
referred to as “hyperbolic” direction finding because we determine the location of the
transmitter using these hyperbolae.

6.2 Using hyperbolae

To create a hyperbola, two receivers are needed, and the time delta between the receive
times at both receivers is measured. If the time delta is zero, then the hyperbola will be a
straight line between the two receivers because every point on this line is the same
distance from both receivers. On the other hand, the hyperbola will be curved if the
difference in the time of arrival at the two receivers is non-zero (Figure 18).

Figure 158 - Creating hyperbolae based on time differences

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 19

 Time Difference of Arrival

If there are three receivers, there will be three

hyperbolae – one between receivers 1 and 2,
another between receivers 2 and 3, and a third
between receivers 1 and 3. The emitter location is
at the intersection of these three curves (Figure 19).
It is possible to have more than three stations, but a
minimum of three stations is needed to obtain an
unambiguous position fix.

6.3 Implementation of TDOA Figure 19 - Emitter location based on TDOA

A TDOA-based direction finding system is made up of a network of interconnected sensors

and a master station. Received signals of interest are digitized, that is, are converted to IQ
data, and then transferred to the master station over a data link. Note that TDOA requires
this data to be very precisely timestamped using a shared clock – usually, this precise
shared clock is provided by GPS. The master station then calculates cross-correlations
using the data from each station, and the results of this calculation provide the time
difference, which is then used to draw the hyperbolae.

6.4 Correlating signals

TDOA requires the difference in the time of arrival (delta t), and this time difference is
determined using correlation. A reference time is defined and correlation is calculated at
various time offsets. The results of the correlation yield a correlogram, whose peak value
corresponds to the time offset that produces the strongest correlation (Figure 20).

Figure 20 - Determining delta t using a TDOA correlogram

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 20

 Time Difference of Arrival

When the signal is wideband, it is usually not difficult to find a clear correlation peak for
use in TDOA. However, if the signal is narrowband or CW (unmodulated), it becomes
much more difficult to find a clear correlation peak in the correlogram (Figure 21). Without
a clear peak to provide difference in time of arrival, it is difficult to generate an accurate
hyperbola. It is important to remember that TDOA-based direction finding system yield
more accurate results when trying to locate wideband signals.

Figure 21 - Ambiguity when calculating delta t for an unmodulated signal

6.5 TDOA sensors

All of the previous direction finding methodologies discussed so far, such as Doppler,
Watson-Watt, or correlative interferometry each required a special type of directional
antenna or antenna array. TDOA, on the other hand, is concerned with the time at which a
signal arrives rather than the direction or angle from which that signal arrives and requires
collection and timestamping of data at multiple points simultaneously. Therefore, TDOA-
based direction finding typically is implemented in the form of purpose-built sensors that
contain a receiver and a non-directional antenna. These sensors can be more or less
permanently fixed in a location or can be movable. Sensors digitize received radio
frequency signals in the form of IQ (in-phase and quadrature) data, and timestamp this
data, most commonly obtaining timing from GPS or another satellite navigation system.
This timestamped data is then sent to the master data station over some type of data link.
Good accuracy over a larger area requires a large number of these sensors, so dedicated
TDOA sensors tend to be relatively inexpensive with lower RF performance than the
receivers used in other direction finding methodologies.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 21

 Time Difference of Arrival

6.6 Location coverage and accuracy

In general, TDOA results are accurate to about several hundred meters, this being both a
function of the methodology and sensors themselves, as well as the target signal type, that
is, broadband versus narrowband. Note too that best accuracy is obtained when the target
is more or less “surrounded” by sensors, but that outside of this area, location accuracy
can be poor. For example, if a target has sensors on all sides, accuracy will usually be
good. On the other hand, if the target is outside of the sensor network, results are often
poor (Figure 22). So often one needs to have some idea of where the target is or will be
before sensors are deployed. This is not a problem when trying to cover a given area, but
limits the usefulness of TDOA in ad hoc direction finding applications.

Figure 22 - Influence of sensor geometry in TDOA

6.7 Practical considerations in TDOA

As shown above, TDOA location accuracy is usually good inside the receiver coverage
area, that is, when the target is “surrounded” by as many sensors as possible. TDOA
accuracy tends to be much better for wideband than narrowband signals, mostly due to the
difficulty in getting a strong correlation peak and thus a time offset for narrowband signals.
Adding additional TDOA sensors to the network offers better accuracy, but at some point
(typically about six sensors), there is no additional improvement and calculation time
increases dramatically. TDOA also offers something called "proximity gain," which means
that a system is more likely to receive weaker signals if there are many sensors in an area
– the odds of one of the sensors being physically “close” to the source is higher. And
finally, some TDOA algorithms can detect and process out the effect of multipath to
various degrees.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 22

 Hybrid Methodologies

7 Hybrid Methodologies
7.1 What is a hybrid methodology?
A recent advancement in direction finding is something called “hybrid methodologies.” One
possible definition of a “hybrid” methodology is something that is actually quite common in
many direction finding applications, namely using an automatic direction finding system to
obtain the general location of a target, e.g. within about a hundred meters and then
switching to manual or hand-held direction finding to narrow down the location to the meter
level. In a sense this is a “hybrid” methodology, but normally what is meant by the term
“hybrid” methodology is the combination of two different automatic direction finding
methodologies: in particular, the combination of an angle of arrival methodology and a time
difference of arrival methodology.

7.2 Combining angle of arrival and time difference of arrival

As we known, there are two different methods of using bearings. The first of these is
homing, which gives a direction or bearing towards the target. This is most often used in
mobile applications where one can “follow the bearing” to the target. The other method of
using bearings is triangulation, where bearings are taken either from multiple locations,
using multiple direction finders, or by moving the direction finder from place to place.
In time difference of arrival, the receivers or stations are usually fixed, and location
information is derived from the differences in the time when the target signal appears at
each station, represented in the form of hyperbolae. Note that in angle of arrival, one could
use a single mobile DF system to take bearings from different locations one after another
and then later combine them. A TDOA based DF system, on the other hand, requires a
minimum of three stations operating *simultaneously* to produce unambiguous results.
A hybrid scenario uses one or more TDOA sensors and one or more angle of arrival
stations (Figure 23). In this example, two TDOA sensors create a hyperbola, with the
target lying somewhere along this line. Rather than using an additional TDOA station to
locate the target, an AOA station generates a bearing, which provides the information
needed to make an unambiguous position determination.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 23

 Hybrid Methodologies

Figure 23 - Hybrid DF using both AOA and TDOA

7.3 Advantages of hybrid methodologies

Additional flexibility can be gained by integrating TDOA and angle of arrival functionality
into a hybrid station. For example, when trying to locate a narrowband source in a
relatively low multipath environment, the stations operate in angle of arrival mode. If trying
to locate a wideband signal source or when operating in a high multipath environment,
switching two or more of these stations to TDOA may yield better results. The benefit of a
hybrid approach is that the operator can try different combinations of stations and
methodologies to see if the results are consistent, or which set of stations yields the
highest accuracy.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 24

 Conclusion

8 Conclusion
All direction finding methodologies have the same goal: determining the physical or
geographical location of a source of radio frequency energy. Angle of arrival based
methodologies create bearings, or directions towards the target, by examining changes in
the received signal's frequency, amplitude, or phase, whereas time difference of arrival
compares the relative receive time of a target signal at geographically separated stations.
The choice of an appropriate direction finding methodology to use for a given application is
largely a function of the target signal's characteristics, such as frequency and modulation,
but is also influenced by the propagation environment (multipath) as well as cost /
complexity. Recent advances in the development of hybrid direction finding methodologies
attempt to overcome some of these restrictions and increase accuracy by using a
combination of methodologies.

1.00 Rohde & Schwarz An Introduction to Direction Finding Methodologies 25

Rohde & Schwarz Regional contact
The Rohde & Schwarz electronics group offers Europe, Africa, Middle East
innovative solutions in the following business fields: +49 89 4129 12345
test and measurement, broadcast and media, secure customersupport@rohde-schwarz.com
communications, cybersecurity, radiomonitoring and
North America
radiolocation. Founded more than 85 years ago, this
1 888 TEST RSA (1 888 837 87 72)
independent company has an extensive sales and customer.support@rsa.rohde-schwarz.com
service network and is present in more than 70
countries. Latin America
+1 410 910 79 88
The electronics group is among the world market
customersupport.la@rohde-schwarz.com
leaders in its established business fields. The
company is headquartered in Munich, Germany. It Asia Pacific
also has regional headquarters in Singapore and +65 65 13 04 88
Columbia, Maryland, USA, to manage its operations customersupport.asia@rohde-schwarz.com
in these regions.
China
+86 800 810 82 28 |+86 400 650 58 96
customersupport.china@rohde-schwarz.com

Sustainable product design

ı Environmental compatibility and eco-footprint
ı Energy efficiency and low emissions
ı Longevity and optimized total cost of ownership

This white paper and the supplied programs may only

be used subject to the conditions of use set forth in
the download area of the Rohde & Schwarz website.
R&S® is a registered trademark of Rohde & Schwarz GmbH & Co.
KG; Trade names are trademarks of the owners.

Rohde & Schwarz GmbH & Co. KG

Mühldorfstraße 15 | 81671 Munich, Germany
Phone + 49 89 4129 - 0 | Fax + 49 89 4129 – 13777

www.rohde-schwarz.com