949

1 INTRODUCTION

In recent years, there has been a dynamic increase in

the use of unmanned aerial vehicles (UAVs),

commonly known as drones, in military operations.

Their role on the battlefield continues to evolve—from

reconnaissance tools that support command and

observation to precise strike platforms with a high

degree of autonomy. Conflicts such as the 2020

Nagorno-Karabakh war and the ongoing war in

Ukraine clearly demonstrate how significantly drones

are reshaping the nature of modern warfare [1], [2]. The

growing availability of drone technology and the

relatively low cost of production have made these

systems increasingly attractive to both state actors and

irregular armed groups.

The application of UAVs that we are investigating

in several scientific research projects is the use of radio

emitters for localization. The method that we use is the

SDF (signal Doppler frequency) method. It has both

advantages and disadvantages. The aim of this article

is to examine the influence of selected parameters on

the accuracy of localization of radio emitters. The

conclusions from the article will allow us to focus on

the most important ones, which have the greatest

impact on the localization errors. They show directions

for further development of the method to improve its

accuracy.

The rest of the paper is organized as follows. Section

2 describes the SDF method and her parameters, that

influence the localization error. Simulation scenarios

and obtained results are described in Section 3. The

conclusions that are included in Section 4 concern the

recommendations and directions for further research

into the SDF method.

Study of Parameters Influencing the Accuracy

of the SDF Method Localization

K. Bednarz, C. Ziółkowski & J. Wojtuń

Military University of Technology, Warsaw, Poland

ABSTRACT: Modern military operations increasingly use unmanned aerial vehicles (UAVs) not only for

observation and reconnaissance, but also for active localization of radio emission sources. One of the methods

used for this purpose is the signal Doppler frequency method (SDF), based on the analysis of the frequency of the

signal received by the moving sensor. The paper presents a theoretical analysis and simulation studies aimed at

determining the effect of selected parameters on the accuracy of emitter localization using the SDF method. In

particular, the factors such as data acquisition time, accuracy of Doppler frequency estimation, carrier frequency

and the velocity of the moving sensor were considered. The aim of the work is to indicate which of these

parameters are crucial for the quality of localization. We formulate conclusions that can support the development

of the resistant to interference and more precise localization systems based on the SDF method. The presented

approach can be used both in real armed conflicts and in work on autonomous electronic reconnaissance systems.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 19

Number 3

September 2025

DOI: 10.12716/1001.19.03.29

950

2 SDF METHOD

2.1 SDF Method principles

The SDF method is a frequency method for locating

radio emitters. The basic parameter on which the

localization algorithms operate is the Doppler

frequency shift (DFS), which appears when the receiver

or transmitter is set in motion [3]. This frequency can

be expressed by the formula:

f f cos



(1)

where: v – speed of change of position between the

signal source and the receiver, f0 – frequency of the

emitted carrier wave, c- the speed of propagation of an

electromagnetic wave in a medium, γ- angle between

the direction of the velocity vector and the direction

determined by the position of the signal source and

receiver.

Assuming that we localize the stationary

transmitter using a sensor placed on a moving

platform, e.g. an unmanned aerial vehicle (UAV), we

can estimate the coordinates

( )

,xy

of the localized

emitter in a two-dimensional plane using the equations

[4]:

( ) ( )

1 1 2 2

t A t t A t

A t A t

−



−

(2)

( ) ( ) ( )

( ) ( )

1 2 1 2

v t t A t A t

A t A t



−

  −



−





(3)

where:

( )

, , , ,

max

A t F t f kf k

F t f c

−

=  = =

fDmax –

maximum Doppler frequency, f0 - carrier frequency of

transmitted signal, z- constant height relative to the

receiver, v- velocity of receiver.

2.2 Parameters affecting the localization error

Analyzing equation (1), we can see that the

fundamental element influencing the Doppler

frequency fD value, apart from the carrier frequency f0

is the angle between the direction of the velocity vector

and the direction determined by the position of the

signal source and receiver γ. How the UAV flight

direction relative to the emitter location will affect the

nature of the Doppler frequency fD changes and the

SDF method location error is presented, among others,

in [5], [6]. Therefore, it is extremely important to select

the correct UAV flight direction, which is worth noting

here. For this reason, in further considerations

including research on other parameters, many

scenarios of UAV location relative to the emitter will be

taken to develop optimal solutions into account.

However, scenarios in which the Doppler frequency is

extremely low or its value is almost constant, will not

be considered.

An analysis of formulas (2) and (3) shows that the

accuracy of localization is influenced by the accuracy

of determination:

− Doppler frequency fD,

− carrier frequency of transmitted signal f0,

− velocity of receiver v.

When spectrum analysis is used to estimate the

Doppler frequency fD, it can be expressed as [7]:

( ) ( )

f t f t f=−

(4)

where: f(t)- instantaneous frequency of received signal.

In this case, the accuracy of the carrier frequency f0

estimation of the emitted signal also affects the

accuracy of the Doppler frequency fD estimation. In the

following article, it is assumed that the Doppler

frequency fD is estimated using methods which do not

need information about the carrier frequency of the

signal. Consequently, carrier frequency estimation

error affects the accuracy of localization only in

formulas (2) and (3).

Formulas (2) and (3) allow for estimating the

coordinates (x,y) of the localized emitter in the case of

measuring the Doppler frequency fD at least at two

moments of time t1 and t2. Using a larger amount of

data, e.g. a ten-element vector, will also affect the

localization error. The fourth analyzed parameter will

therefore be:

− the acquisition time t_A, reflecting the amount of

data (Doppler frequency shifts fD) used to determine

the coordinates (x,y) of the emitter.

As with the UAV velocity v, information about the

UAV location may also be subject to some error. The

SDF method determines the position relative to the

sensor position, and therefore it will also affect the

error in the radio emission location. For this reason, it

was decided to also examine the influence of

− the UAV’s positioning accuracy.

In the case of performing localization procedures in

real time, we rely only on a part of the Doppler curve,

and we cannot select an arbitrary acquisition time tA.

However, it is necessary to determine when to perform

localization procedures so that the result is not

burdened with too large localization error ∆r. The

solution to this problem may be the last parameter

studied in the publication, called:

− the range of Doppler frequency changes ∆fD.

3 SIMULATION STUDIES FOR TESTING THE

IMPACT OF SELECTED PARAMETERS ON

LOCLIZATION ACCURACY

3.1 Main assumptions for all scenarios

In order to investigate the influence of the parameters

listed in Section 2.2 on the localization accuracy of the

SDF method, a research scenario was adopted in which

the localization sensor was mounted on the UAV and

moved along the OX axis of the coordinate system

shown in Figure 1.

951

Figure 1. Spatial scenario for simulation studies.

Based on the arrangement of elements shown in

Figure 1, the following assumptions were made:

− the emitter is located at point x0, y0, z0 of the

coordinate system,

− the localized emitter transmits a harmonic signal at

the carrier frequency f0,

− the carrier frequency f0 of the emitted signal is

determined by the sensor mounted on the non-

moving UAV No. 2, which is placed at coordinates

(xv, yv, zv)=(0,0,0)km,

− the Doppler frequency fD is determined based on the

signal samples received by both sensors,

− the carrier frequency f0 estimation error of the signal

does not affect the Doppler frequency fD estimation

error, and consequently, carrier frequency f0

estimation error affects the accuracy of localization

only in formulas (2) and (3),

− UAV No. 1 moves along the OX axis over a distance

of S km with velocity v, at subsequent moments of

time its position coordinates are (xv, yv, zv)

=(vt,0,0)km,

− the localization sensor estimates parameters (f0, fD,

v) every 1 s,

− to determine the emitter coordinates (x,y) data

vectors (f0,fD,v) of length equal to the acquisition

time tA are taken,

− the localization error ∆r is determined according to

formula

 = − + −

( ) ( )r x x y y

where x(t) and y(t) are the coordinates estimated

based on formulas (2) and (3).

3.2 The influence of the acquisition time

A similar analysis of this parameter was performed in

[8]. The authors of the article analyzed the acquisition

time tA for recorded signals burdened with large

oscillations. In this chapter, we would like to present

the trend of the localization error ∆r in the case of the

selection of the acquisition time tA for ideal conditions.

3.2.1 Scenarios assumptions

The study was conducted taking the assumptions

from Section 3.1 into account. Additionally, the

following assumptions were made:

− x0,y0,z0={(10,40,-0.1),(10,30,-0.1),(10,20,-0.1),

(10,10,-0.1),(10,5,-0.1),(10,2,-0.1),

(10,1,-0.1),

(5,20,-0.1),(5,15,-0.1),(5,10,-0.1),

(5,5,-0.1),(5,2.5,-0.1),(5,1.25,-0.1),

(5,0.5,-0.1),

(1.2,4.8,-0.1),(1.2,3.6,-0.1),(1.2,2.4,-0.1),

(1.2,1.2,-0.1),(1.2,0.6,-0.1),(1.2,0.2,-0.1),

(1.2,0.1,-0.1)} km,

− f0={300,600,900,1200,1500,1800,2100,2400,2700,

3000} MHz,

− v=15 m/s,

− the selection of acquisition time tA and its maximum

value results from the selected scenarios. When the

path S=2∙x0 is 20 kilometers, at a speed of 15 m/s we

need 1334 seconds to complete the entire route.

When S=10 kilometers we need 667 seconds and for

S=2400 meters it is 161 seconds. It was therefore

decided to examine the acquisition time tA values

from 2 seconds to the total number of seconds

needed to cover the entire route tAmax, with a step of

1 s for all cases,

− the emitter position coordinates are determined for

an analysis window of the acquisition time tA

length. Then, if the coordinate determination

counter k is lower than the maximum number of

localization procedure repetitions K defined as

= − +

max

K t t

(6)

the analysis window is shifted by one second and

the coordinates are determined again. This

procedure is repeated K times. For this reason, the

first location result is obtained after a time equal to

the acquisition time tA.

− for each value of the acquisition time tA, the mean

value μ∆r and deviation σ∆r of the localization error

∆r is determined according to formulas



Δ,

μr

(7)

=−

−



ΔΔ

Δ,

r k r

σ r μ

(8)

where k=1,2,…,K, ∆rk is a localization error defined

by Equation (5) for the kth estimation of emitter

coordinates (x(t),y(t)).

3.2.2 Results

Table 1 presents the mean value μ∆r and standard

deviation σ∆r of the localization error for each value of

the acquisition time tA. Due to the different number of

seconds needed to cover the entire route tAmax

depending on the length of the path S, the acquisition

time tA was expressed in percentage as:

 

% 100

Amax

=

(9)

The course of the mean localization error as a

function of acquisition time tA for different path lengths

is presented in Figure 2.

952

Table 1. The mean value μ∆r and standard deviation σ∆r

of the localization error for different path length S and

different value of the acquisition time tA.

S=20 km

S=10 km

S=2.4 km

tA [%]

Localization error [m]

μ∆r

σ∆r

μ∆r

σ∆r

μ∆r

σ∆r

1.471

3.356

0.684

1.612

0.160

0.321

1.114

2.316

0.524

1.119

0.124

0.225

0.836

1.499

0.396

0.723

0.096

0.148

0.632

0.881

0.304

0.428

0.075

0.090

0.500

0.454

0.243

0.222

0.062

0.051

0.440

0.231

0.216

0.116

0.056

0.031

0.442

0.198

0.216

0.101

0.056

0.027

0.466

0.213

0.228

0.108

0.059

0.029

0.491

0.220

0.240

0.111

0.063

0.031

100

0.530

0.245

0.258

0.123

0.069

0.037

Figure 2. The mean value of localization error μ∆r as a

function of acquisition time tA for different path lengths S.

Analyzing the above results, we can see that the

lowest mean localization error μ∆r is obtained in the

case when the acquisition time tA is about 65% of all

seconds needed to cover the entire route tAmax.

However, this value is difficult to achieve, especially in

the case of scenarios where there are more than 1300 of

them. The measurement time to the first localization

would reach more than 10 minutes. It can also be seen

that with the increase in the length of the measurement

path S, the difference between the localization errors

for individual acquisition times tA increases. However,

the trend itself is identical. In ideal simulation

conditions, we can see, however, that the selection of

the acquisition time tA does not have a critical

significance in the localization error.

3.3 The influence of the range of Doppler frequency

changes

As mentioned in Section 3.2, the accuracy of

localization is influenced by the acquisition time tA of

the signal, i.e. the number of fD values taken to calculate

the emitter coordinates (x,y). In general, the lowest

localization error was obtained when the acquisition

time tA was about 60% of the entire Doppler curve. In

real conditions, we do not always have the possibility

to collect data that will allow for the analysis of the full

Doppler curve and localization procedures must be

performed on its part. In such a case, a decision must

be made when to start determining the coordinates so

as not to burden the result with too large a localization

error ∆r. For this purpose, another simulation study

was conducted, in which it was decided to analyze the

influence of the range of Doppler frequency changes

BfD on the localization error ∆r. For assume that this

parameter has a constant value, the acquisition time tA

is changed dynamically by appropriately adding and

subtracting values from the fD vector, to maintain

a constant, previously determined the range of

Doppler frequency changes BfD.

3.3.1 Scenarios assumptions

The study was conducted taking the assumptions

from Section 3.1 into account. Additionally, the

following assumptions were made:

− x0,y0,z0={(10,10,-0.1),(10,5,-0.1),(10,2,-0.1),

(10,1,-0.1),(1.2,1.2,-0.1),(1.2,0.6,-0.1)

(1.2,0.2,-0.1),(1.2,0.1,-0.1)} km,

− f0={300,600,900,1200,1500,1800,2100,2400,2700,

3000} MHz,

− v={15,20,30} m/s,

− the range of Doppler frequency changes was

selected so that each scenario could meet the

requirements for its value, it was therefore decided

to test all possible BfD values in the range from 0.5

to 20 Hz with a step of 0.5 Hz,

− the acquisition time tA is selected taking the range of

Doppler frequency changes BfD into account,

− for each value of the range of Doppler frequency

changes BfD, the mean value μ∆r and deviation σ∆r

of the localization error ∆r is determined according

to formula (7) and (8).

3.3.2 Results

Figure 3 presents courses of the mean value μ∆r and

standard deviation σ∆r of the localization error as a

function of the range of Doppler frequency changes

BfD.

Figure 3. The mean value μ∆r and standard deviation σ∆r of

localization error as a function of the range of Doppler

frequency changes BfD.

Analyzing the obtained results, it can be noticed

that with small values of the range of Doppler

frequency changes BfD, the mean localization error μ∆r

can reach almost 3 meters. By increasing the value of

the tested parameter, localization error begins to

decrease quickly. Assuming that the localization

procedure is to be carried out as quickly as possible and

the error cannot exceed 0.5 meters, localization can be

done on data for which the range of Doppler frequency

changes BfD equals 5 Hz. Further reduction of the value

of this parameter obviously causes a decrease in the

localization error and for BfD=20 Hz, the mean

localization error value μ∆r=0.278 m. However, the

parameter study was performed for ideal conditions

953

and in the case of real conditions, where the Doppler

frequency will oscillate, among others due to the

frequency stability of the SDR platform used [9], it will

be necessary to previously determine the value of this

parameter for the target solution.

3.4 The influence of the Doppler frequency estimation

error

The approximate analysis was carried out in [10]. The

ideal Doppler frequency courses as a function of time

were obtained using the following equation

( )

( )( )

2 2 2

0 0 0

, ,

x vt

f t k f

x vt k y z





−



−



− + − +



where x=(x0,y0,z0).

In our consideration it was assumed, that the

estimated fD values oscillate and the oscillations follow

a normal distribution N(μfD, σfD).

3.4.1 Scenarios assumptions

The study was conducted taking into account the

assumptions from section 3.1. Additionally, the

following assumptions were made:

− x0,y0,z0=(1.5,1.0,-0.1) km,

− Doppler frequency fD oscillations have a normal

distribution N(μfD, σfD) = N(0, 0.001), N(0, 0.01), N(0,

0.1), N(0, 1), N(0, 10),

− f0=1200 MHz,

− v=15 m/s ,

− the acquisition time tA=30 s,

− the mean value μ∆r and standard deviation σ∆r of the

localization error ∆r is determined according to

formula (7) and (8).

3.4.2 Results

Figure 4 shows the Doppler curves for all the

scenarios studied. Table 2 and Figure 5 show the

localization error for the given scenarios.

Figure 4. The Doppler curves for all the scenarios.

Figure 5. Localization error as a function of time for all

distributions of Doppler frequency f_D oscillations.

Table 2. The mean value μ∆r and standard deviation σ∆r of

the localization error for different distributions of Doppler

frequency fD oscillations.

Localization error [m]

N(μfD, σfD) [Hz]

μ∆r

σ∆r

N(0, 10)

857.08

1471.41

N(0, 1)

38.06

31.82

N(0, 0.1)

5.07

4.30

N(0, 0.01)

0.58

0.54

N(0, 0.001)

0.05

Analyzing the obtained results, it can be seen that

even small oscillations of the estimated Doppler

frequency fD lead to significant errors. In the case of

σfD=1 Hz, a satisfactory localization error can be

obtained. As can be seen, σfD=10 Hz disqualifies the

SDF method from practical use without prior filtering

of Doppler frequency curves.

3.5 The influence of the carrier frequency estimation error

As mentioned in Section 2.2 in the following article, it

is assumed that the Doppler frequency fD is estimated

using methods which do not need information about

the carrier frequency of the signal. Consequently,

carrier frequency estimation error affects the accuracy

of localization only in formulas (2) and (3).

3.5.1 Scenarios assumptions

The study was conducted taking the assumptions

from Section 3.1 into account. Additionally, the

following assumptions were made:

− x0,y0,z0=(1.5,1.0,-0.1) km,

− Doppler frequency fD is estimated without error,

− carrier frequency of transmitted signal

f0=1200 MHz,

− v=15 m/s ,

− the acquisition time tA=30 s,

− carrier frequency f0 is estimated with error

∆f0={0.01,0.1,1,10,100} MHz which is added to real

value of carrier frequency of transmitted signal f0,

− the mean value μ∆r and standard deviation σ∆r of the

localization error ∆r is determined according to

formula (7) and (8).

3.5.2 Results

Table 3 and Figure 6 show the localization error for

the given scenarios.

954

Figure 6. Localization error as a function of time for different

estimated carrier frequency (f_0=1200 MHz) .

Table 3. The mean value μ∆r and standard deviation σ∆r

of the localization error for different errors of carrier

frequency estimation.

Localization error [m]

Estimated f0 [MHz]

μ∆r

σ∆r

1300.00

238.03

145.09

1210.00

24.30

14.97

1201.00

2.39

1.48

1200.10

0.20

0.12

1200.01

0.02

Analyzing the above results, we can see that even a

carrier frequency estimation error ∆f0 of 1 and 10 MHz

does not disqualify the SDF method from use. The

localization error does not exceed 2.5 and 25 meters in

this case, respectively. Only at ∆f0=100 MHz does the

localization error reach hundreds of meters, which

would not allow for correct localization for a given

scenario.

3.6 The influence of the UAV’s velocity fluctuations

3.6.1 Scenarios assumptions

The study was conducted taking the assumptions

from Section 3.1 into account. Additionally, the

following assumptions were made:

− x0,y0,z0=(1.5,1.0,-0.1) km,

− Doppler frequency fD is estimated without error,

− carrier frequency of transmitted signal f0=1200 MHz

and is estimated without error,

− the acquisition time tA=30 s,

− velocity v is estimated. Oscillations have a normal

distribution N(μv, σv) = N(0, 0.001), N(0, 0.01), N(0,

0.1), N(0, 1) and they are added to constant velocity

v=15 m/s,

− the mean value μ∆r and standard deviation σ∆r of the

localization error ∆r is determined according to

formula (7) and (8).

3.6.2 Results

Figure 7 shows the UAV velocity courses as a

function of time for all the scenarios studied. Table 4

and Figure 8 show the localization error for the given

scenarios.

Figure 7. The UAV velocity courses as a function of time for

all the scenarios studied.

Figure 8. Localization error as a function of time for all

distributions of velocity v oscillations.

Table 4. The mean value μ∆r and standard deviation σ∆r of

the localization error for different velocity v estimation

error.

Localization error [m]

N(μv, σv) [m/s]

μ∆r

σ∆r

N(0,1)

229.59

214.56

N(0, 0.1)

22.59

19.52

N(0, 0.01)

2.42

2.33

N(0, 0.001)

0.22

0.21

Analyzing the above results, we can see that an

error in determining the UAV speed of about 1 m/s

results in localization errors of several hundred meters.

Therefore, it is necessary to estimate this parameter as

accurately as possible. The distribution of the UAV

speed estimation error for which the localization error

is acceptable is N(0,0.1).

3.7 The influence of the UAV’s positioning accuracy

3.7.1 Scenarios assumptions

The study was conducted taking the assumptions

from Section 3.1 into account. Additionally, the

following assumptions were made:

− x0,y0,z0=(1.5,1.0,-0.1) km,

− Doppler frequency fD is estimated without error,

− carrier frequency of transmitted signal f0=1200 MHz

and is estimated without error,

− the acquisition time tA=30 s,

− the determination of its own position by the UAV is

performed with an error whose values have a

normal distribution N(μP, σP) = N(0, 0.1), N(0, 1),

N(0, 2), N(0, 5), N(0, 10) m,

955

− the velocity is obtained from the UAV and is equal

to v=15 m/s

− the mean value μ∆r and standard deviation σ∆r of the

localization error ∆r is determined according to

formula (7) and (8).

3.7.2 Results

Figure 9 shows an example of a UAV route used in

research with and without UAV location error. Table 5

show the localization error for the given scenarios.

Figure 9. Example of a UAV route used in research with and

without UAV location error.

Table 5. The mean value μ∆r and standard deviation σ∆r of

the localization error for different UAV location errors.

Localization error [m]

N(μP, σP) [m]

μ∆r

σ∆r

N(0, 0.1)

0.41

0.25

N(0, 1)

3.64

2.72

N(0, 2)

7.39

4.98

N(0, 5)

18.33

13.89

N(0, 10)

37.49

26.61

Analyzing the above results, we can see that an

error in determining the UAV location also has a

significant impact on accuracy of radio emissions

localization. However, it does not affect that

significantly as, for example, the Doppler frequency fD.

Considering the accuracy of available GNSS receivers,

whose worst-case location error is about 3-5 meters,

will still allow for obtaining effective location for which

the location error of the SDF method will increase by

only a few meters.

4 CONCLUSIONS

The aim of the article was to examine the influence of

parameters on the accuracy of localization of emission

sources using the SDF method. For this purpose,

a theoretical analysis of the SDF method and a series of

simulation studies were conducted.

Analyzing the obtained results, it can be seen that

in ideal conditions and taking all research assumptions

into account, the carrier frequency f0 estimation error

has the smallest impact on the accuracy of localization.

In its case, even an estimation error reaching tens of

megahertz allows for effective determination of the

position of the signal source. The selection of the

acquisition time tA has a slightly greater influence. In

extreme cases, for ideal conditions, its change allows

for a reduction of the localization error ∆r by several

meters, which is not a significant value. The selection

of this parameter may become more important in the

case of an additional error in the estimation of the

Doppler frequency fD. However, the analysis

conducted in the article concerned ideal conditions and

only the change of the acquisition time tA value.

A much greater influence in the case of localization is

the precise determination of the velocity of the

localization sensor carrier, in our case UAV. In this

case, an incorrect determination of the speed differing

from the actual value by only 1 m/s can result in

localization errors reaching hundreds of meters for the

analyzed scenario. Also, the influence of the UAV’s

positioning accuracy is important in SDF method

location error. The scenarios tested show that the use

of commercial GNSS receivers with an error of several

meters should still allow for effective localization, and

the impact on the SDF method location error should be

of the order of several meters. However, the greatest

impact on the localization error can be observed in the

case of the Doppler frequency fD estimation. In the

assumed scenario, its value varied in the range from

about 60 Hz to -60 Hz. Errors in the estimation of this

parameter reaching tens of hertz completely disqualify

the method from practical use. Frequency estimation

with an error of single hertz, on the other hand,

translates into a localization error of several dozen

meters, which, however, for a given scenario allows for

the correct localization of the signal source.

To sum up, in the case of the SDF method, the

greatest impact is played by the Doppler frequency fD

estimation. Its determination is also the biggest

problem in the case of practical implementation,

especially when we are dealing with modulated signals

and non-line-of-sight (NLOS) conditions. Future

research will therefore focus on the most faithful

recovery of the Doppler frequency fD value using

different frequency estimators.

ACKNOWLEDGMENT

This work was developed within the framework of the grant

no. UGB/22-059/2025/WAT, sponsored by the Military

University of Technology (WAT), Poland.

ABBREVIATIONS

DFS Doppler frequency shift

GNSS global navigation satellite system

NLOS non-line-of-sight

SDF signal Doppler frequency

UAV unmanned aerial vehicle

REFERENCES

[1] “Drones in the Nagorno-Karabakh War: Analyzing the

Data,” Military Strategy Magazine. Accessed: Apr. 06,

2025. [Online]. Available:

https://www.militarystrategymagazine.com/article/dron

es-in-the-nagorno-karabakh-war-analyzing-the-data/

[2] N. J. Bradford, “Drones Over Ukraine: Commercial

Technologies in Combat”.

[3] C. Ziółkowski, J. Rafa, and J. M. Kelner, “Lokalizacja

źródeł fal radiowych na podstawie sygnałów

odbieranych przez ruchomy odbiornik pomiarowy,”

Biuletyn Wojskowej Akademii Technicznej, vol. Vol. 55,

no. nr sp., pp. 67–82, 2006.

956

[4] J. M. Kelner and C. Ziółkowski, “Adaptacja technologii

SDF w procedurach lokalizacji i nawigacji,” presented at

the Konferencja Naukowo-Techniczna Systemy

Rozpoznania i Walki Radioelektronicznej, Ołtarzew,

2016.

[5] J. Rafa and C. Ziółkowski, “Influence of transmitter

motion on received signal parameters – Analysis of the

Doppler effect,” Wave Motion, vol. 45, no. 3, pp. 178–190,

Jan. 2008, doi: 10.1016/j.wavemoti.2007.05.003.

[6] J. Kelner, P. Gajewski, and C. Ziółkowski, “Spatial

localization of radio wave emission sources using SDF

technology,” 2012, pp. 367–375.

[7] J. M. Kelner, C. Ziółkowski, and P. Marszałek, “Influence

of the frequency stability on the emitter position in SDF

method,” in 2016 International Conference on Military

Communications and Information Systems (ICMCIS),

May 2016, pp. 1–6. doi: 10.1109/ICMCIS.2016.7496554.

[8] R. Szczepanik and J. Kelner, Two-Stage Overlapping

Algorithm for Signal Doppler Frequency Location

Method. 2023. doi:

10.23919/SPSympo57300.2023.10302723.

[9] K. Bednarz, J. Wojtuń, J. M. Kelner, and K. Różyc,

“Frequency Instability Impact of Low-Cost SDRs on

Doppler-Based Localization Accuracy,” sensors, Feb.

2024, doi: https://doi.org/10.3390/s24041053.

[10] C. Ziółkowski, J. Rafa, and J. M. Kelner, “Przestrzenno-

częstotliwościowe uwarunkowania lokalizacji źródeł fal

radiowych wykorzystującej efekt Dopplera,” Biuletyn

Wojskowej Akademii Technicznej, vol. Vol. 56, no. nr 3,

pp. 7–20, 2007.