Mode-S Radar Interrogation Algorithm Design for Dense Air Traffic Environment

Abstract

The increasing trend in air traffic density will continue in the near future with the addition of different aerial vehicles. Before the Mode-S protocol, Mode A and Mode C were in use; however, the Mode A/C configuration was only usable in sparsely dense air traffic. One of the useful features of Mode-S is the ability of probabilistic interrogation. However, there has not yet been a sophisticated algorithm for many close aircraft. Considering a futuristic air environment with a swarm of drones and airbuses equipped with transponders, we utilized the probabilistic interrogation feature of Mode-S and designed an algorithm. The proposed algorithm is able to collect close aircraft information in a relatively short time. There has also been created a high-level Mode-S uplink and downlink communication simulator in order to exchange all-call communication and record the algorithm’s performance in terms of time and number of interrogations sent.

Full text

460 A. ONCU, A. G. AYDIN, Y. ERDOGAN, ET AL., MODE-S RADAR INTERROGATION ALGORITHM DESIGN …
DOI: 10.13164/re.2022.0460
Mode-S Radar Interrogation Algorithm Design
for Dense Air Traffic Environment
Ahmet ONCU, Ahmet Gunhan AYDIN, Yunus Emre ERDOGAN, Artun AKDOGAN
Dept. of Electrical and Electronics Engineering, Bogazici University, 34342 Bebek Istanbul, Turkey
{ahmet.oncu, gunhan.aydin, yunus.erdogan, artun.akdogan}@boun.edu.tr
Submitted May 16, 2022 / Accepted September 21, 2022 / Online first October 7, 2022
Abstract. The increasing trend in air traffic density will
continue in the near future with the addition of different
aerial vehicles. Before the Mode-S protocol, Mode A and
Mode C were in use; however, the Mode A/C configuration
was only usable in sparsely dense air traffic. One of the
useful features of Mode-S is the ability of probabilistic
interrogation. However, there has not yet been a sophisti-
cated algorithm for many close aircraft. Considering
a futuristic air environment with a swarm of drones and
airbuses equipped with transponders, we utilized the prob-
abilistic interrogation feature of Mode-S and designed
an algorithm. The proposed algorithm is able to collect
close aircraft information in a relatively short time. There
has also been created a high-level Mode-S uplink and
downlink communication simulator in order to exchange all-
call communication and record the algorithm’s perfor-
mance in terms of time and number of interrogations sent.
Keywords
Garbling, lockout, Mode-S, stochastic interrogation
1. Introduction
The Mode-S (Selective) civilian communication sys-
tem being used in secondary surveillance radars with
1030 MHz interrogation and 1090 MHz reply frequencies
was first introduced in the 1970s by MIT Lincoln Laboratory
and improved through the 1980s and 1990s. Prior to the
Mode-S protocol, Mode A and Mode C were in use; how-
ever, this configuration was only usable in sparsely dense air
traffic [1]. The reason is that all SSR (secondary surveillance
radar) replies stack in the same frequency band; hence all
aircraft signals being detected in the same antenna beam
overlap with each other, and this phenomenon is known as
garbling in general. When the 1090 MHz band is shared with
both the Mode-S, ADS-B, and Mode A/C systems, there
occurs FRUIT (false replies unsynchronized in time) which
is co-channel interference due to dense environments [2].
Mode A and Mode C give information about interro-
gated aircraft’s squawk code (4 octal digits) and altitude,
respectively [3]. Besides mentioned traditional SSR abili-
ties, Mode-S provides enhanced surveillance and datalink to
ground control stations to be able to monitor aircraft with
more capability.
Even operations with Mode-S systems cannot avoid
garbling problems in the dense air traffic regions in terms of
traffic since there occur close encounters between different
aircraft within a close range, which is five nautical miles
usual standard for horizontal separation, and around
2000 feet (~0.38 NM) for vertical separation [4]. These
close scenarios can be caused by different sources such as
lack of capacity of the coverage, weather conditions, mili-
tary activities, airline decisions, etc. [5]. Another reason for
having a highly dense air environment with close encounters
is that there is an immense increase in air traffic demand and
the number of aircraft registered. According to an annual re-
port by EUROCONTROL, total flights will increase by
an average of 1.9% per year over the next years, reaching
a total of 11.6 million flights in 2023 [6]. Furthermore, fore-
casted IFR (instrument flight rules) movements per traffic
zone tell that there is an upward trend in all states of the eu-
rozone states in terms of air traffic density through 2024 [7].
Another reason for seeing many more aircraft on the skyline
is that the cost of air traffic management is relatively discon-
nected from the air traffic, and this allows for the introduc-
tion of more registered aircraft [8]. Speaking of registered
aircraft, the EU continues to develop the controlled inclusion
of unmanned aerial vehicles in the air environment with its
U-space project. This project offers UAS (unmanned aircraft
systems) traffic management (UTM) to integrate UAS into
air traffic management (ATM) [9].
There has been numerous research parallel to the in-
creasing number of transponder-equipped drones for the last
decade to provide airspace safety [10–12]. The increase in
registered aircraft, including unmanned aerial vehicles,
forces authorities to revise regulations such as RPAS (re-
motely piloted aircraft systems) [13]. The number of drones
in urban and rural airspaces will also amplify when passen-
ger transportation [14], package delivery [15], agricultural
surveillance and air cargo [16] are taken into account. From
a military standpoint, border monitoring, landmine detec-
tion, logistic security, and military mission delivery require
more drone injection into the airspace [17], [18].

RADIOENGINEERING, VOL. 31, NO. 4, DECEMBER 2022 461
Therefore, the future of air traffic will be crowded with
UAS inclusion, and this requires new techniques and utilities
in communication systems in order to detect aircraft in the
dense air traffic regions rapidly.
At this stage, in order to reduce the time taken to detect
aircraft in garbled replies, we come up with a probabilistic
interrogation algorithm using already existing Mode-S fea-
tures. In our approach, we use all-call interrogations with
different probabilities in the uplink, and we interrogate with
1, 0.5, 0.25, 0.125, and 0.0625 probabilities according to our
adaptive algorithm. By adaptively interrogating, a ground
station operator can gather ICAO information of aircraft
relatively fast in the dense garbled region from all-call
replies. A different method can be interleaving scheduling
for Mode-S transactions [19]. Inserting additional waiting
times can diminish the heavy load of the dense air traffic in
the all-call interrogation process. However, additional time
and energy trade-offs appear as well.
Before proceeding into the proposed stochastic interro-
gation algorithm for air traffic control radar systems, the
basic structures and fundamental formats of Mode-S sys-
tems are presented in this section.
1.1 Mode-S Interrogations
There are two types of Mode-S interrogations in terms
of data length, short and long interrogations. Uplink interro-
gations operating at 1030 MHz frequency are sent by
a DPSK (differential phase-shift keying) modulation
scheme. A Mode-S interrogation contains a side lob suppres-
sion pulse (P2) after the initial P1 pulse and additionally a 56
(short) or 112 (long) bit data block. The first two pulses, P1
and P2, last 0.8 µs with a 2.0 µs interval between them [20].
The uplink Mode-S interrogation pulse sequence can be seen
in Fig. 1.
An aircraft reply to the interrogation is modulated via
pulse position modulation at 1090 MHz frequency, and the
pulse sequence contains an 8 µs long preamble with 56
(short) or 112 (long) µs long data block. A 16-bit long pre-
amble is fixed and always “1010000101000000” [21].
An air traffic control radar beacon system (ATCRBS)
consisting of a rotating antenna and transponders typically
sends an all-call interrogation around its environment. Re-
plies to these all-call interrogations include the aircraft’s
identity and the altitude, as usual. In traditional Mode A/C
protocols, there are two pulses, namely P1 and P3, separated
Fig. 1. Mode-S uplink pulse sequence [3].
by 8 or 21 µs according to their length. A P4 pulse lasting
1.6 µs is added to distinguish Mode-S all-call interrogations
from Mode A/C ones [22].
1.2 Mode-S Formats
There are many uplink and downlink formats for
Mode-S. The most commons are UF/DF 4, 5, 11, 20, and 21,
as shown in Tab. 1.
UF/DF 20 and 21 contain Comm-B messages other
than the payload, and there are different types of messages
in Comm-B selection. The options can be depicted in Fig. 2.
UF/DF
Bits
Uplink
Downlink
0 56
Short air-to-air
surveillance
(ACAS)
Short air-to-air
surveillance
(ACAS)
4
56
Altitude Request
Altitude Reply
5
56
Identity Request
Identity Reply
11 56 Mode-S All-Call
Mode-S All-Call
reply
16 112
Long air-to-air
surveillance
(ACAS)
Long air-to-air
surveillance
(ACAS)
17
112
-
Extended Squitter
18 112 -
Extended
Squitter/non-
transponder
19 112 -
Military extended
squitter
20 112
Altitude request via
Comm-A
Altitude reply via
Comm-B
21 112
Identity request via
Comm-A
Identity reply via
Comm-B
24
112
Comm-C (ELM)
Comm-D (ELM)
Tab. 1. UF/DF formats of Mode-S [20].
Fig. 2. Comm-B message types. ELS: Elementary Surveil-
lance, EHS: Enhanced Surveillance, MRAR: Meteoro-
logical Routine Air Report, MHR: Meteorological Haz-
ard Report [20].

462 A. ONCU, A. G. AYDIN, Y. ERDOGAN, ET AL., MODE-S RADAR INTERROGATION ALGORITHM DESIGN …
Comm-B messages allow the ground station to gather
specialized information other than position. More common
BDS formats are BDS-10, 17, and 20, as they correspond to
elementary surveillance. The availability and number of re-
ceived Comm-B messages depend heavily on the density of
air traffic and the rate of interrogations [23].
1.3 Selectivity and Garbling
Mode-S transponders have 255 different 56-bit data
storage caches, which are called BDS (Comm-B data selec-
tor) registers. BDS registers content is updated periodically
by a flight management system. Mode-S interrogators have
two main tasks, one being to detect the replying aircraft and
another being to extract more information from a detected
aircraft [24], [25]. Using a unique ICAO (24-bit aircraft ad-
dress) and getting all necessary information from the mini-
mum number of replies reduce the interrogation density per
aircraft. When the ground station is able to detect an aircraft
with at most two interrogations, this results in avoiding
Mode A/C overlaps, and hence selective interrogations
could be further enabled.
Every region has a lockout map in an SSR Mode-S net-
work. Lockout command is given by a ground station to a se-
lected aircraft. The selected aircraft stops responding to
other calls when it is in a lockout state [26]. The lockout map
determines the area where the ground station locked up
an aircraft. Generating a lockout map and managing these
maps with multiple cities are challenging tasks since the
number of locked-up aircraft is constantly changing [27], [28].
All-call interrogations are used for the initial commu-
nication when an aircraft enters the radar coverage. All-call
interrogation signals don’t have a specific destination; hence
all non-locked-up aircraft have to respond to this call.
A ground station periodically radiates all call interrogations
with a maximum rate of 250 Hz [29]. After receiving a reply
from an aircraft, the ground station extracts the specific air-
craft’s rough position and ICAO address. The next step is to
follow the aircraft and interrogate it with “roll call.” Roll-
call interrogations are aircraft-specific interrogations once it
is in a lockout state. This methodology helps ground stations
to prevent unnecessary all-calls and reduce garbling effects.
The proposed Mode-S interrogation algorithm comes
in handy at this stage since collecting all the all-call replies
from aircraft in the dense region is an issue due to interro-
gating with a hundred percent probability.
2. Stochastic Interrogation Algorithm
The algorithm always works better in the dense air traf-
fic regions than the current Mode-S radar uplink interroga-
tion scheme, as the perception is that aircraft will not be
closer than some limited distances. If any close encounter
happens, then this situation will be short, and garbling ef-
fects will be temporary. However, the consideration in the
developed algorithm is that the air traffic density will in-
crease in the future, and there could be close encounters in
the air with the addition of different aerial vehicles such as
application-specific drones, airbuses, etc. By this logic, the
current mechanism cannot deal with highly dense air envi-
ronments as the number of aircraft increases. One possible
solution to this futuristic dense air traffic problem in civilian
aviation is to interrogate with constant probabilities such as
0.5, 0.25, 0.125, and 0.0625 as Mode-Selective uplink com-
munication allows [21]. In this approach, constantly interro-
gating with 0.5 uplink requests would require too many in-
terrogations while trying to detect all aircraft if the number
of aircraft is too many. On the other hand, interrogating with
0.0625 is inefficient if there are not many aircraft in the tar-
get region. However, in dense environments, after detecting
a number of aircraft by interrogating with a lower probabil-
ity, we can fasten the detection process by increasing the
probability contained in the interrogation. When we are get-
ting a lot of garbling, we can decrease the probability of in-
tercept. We propose an algorithm with these adaptive
changes in such situations that will work much more effi-
ciently than the static interrogation method. In our proposed
method, we can estimate the density, low or high, of the re-
maining aircraft by their overall responses for one or more,
depending on the control length and interrogations. For in-
stance, if we are interrogating with 0.0625 and getting no
replies, we should interrogate with a higher probability. In
this algorithm, we are also using the feature of Mode-S;
when aircraft is locked-out, it will not reply to interrogations
until a particular time passes. Furthermore, our proposed al-
gorithm will decrease the need for operator intervention.
The stochastic interrogation working principle is like
this when the probability of intercept contained in the
interrogation is p.
An aircraft replies with probability p.
An aircraft does not reply with probability 1 – p.
Theoretical calculations of static interrogations to
detect all aircraft in the dense region can be seen by using
binomial distributions with the below definitions:
PDETECTION = Probability of only 1 aircraft replying,
PNO-RESPONSE = Probability of getting zero responses,
PGARBLING = Probability of collision due to multiple replies,
( )
1
DETECTION
1
N
P p pN
−
= −
, (1)
()
NO-RESPONSE
1
N
Pp= −
, (2)
GARBLING DETECTION NO-RESPONSE
1 P PP=−−
, (3)
() ()
1
GARBLING 1 1 1
N
P p pN p
−
=−− +−

. (4)
Here, N is the number of aircraft in the dense region. The
number of all-call interrogations with a static p probability
of collecting all all-call replies from aircraft in the dense area
becomes MDETECTION (p).
( ) ( )
DETECTION 1
11
1
N
k
k
p
Mp
ppk
=

−
= −⋅

∑
. (5)

RADIOENGINEERING, VOL. 31, NO. 4, DECEMBER 2022 463
As can be seen, the total number of interrogations
steeply increases as N grows very large. When N is small,
the minimum number of required interrogations is best with
0.5 static probability. Considering both small and large N
cases, there needs to be an adaptive approach to utilize the
best performances of all static probabilities from 0.5 to
0.0625. Therefore, an adaptive algorithm is required to solve
this futuristic dense air traffic interrogation issue with no
hardware change. Here, the pseudocode of the developed al-
gorithm can be seen below.
Algorithm: Adaptive Probabilistic Interrogation
start procedure
Step 1 Interrogate with P = 1
1 if number_of_aircraft_replied == 0 then
2 return
3 else if number_of_aircraft_replied == 1 then
4 repeat Step 1
5 else if number_of_aircraft_replied > 1 then
6 go Step 2
7 end-if
Step 2 Interrogate with P= 1/2
8 if (number_of_aircraft_replied == 0 or
number_of_aircraft_replied == 1) then
9 go Step 1
10 else if number_of_aircraft_replied > 1 then
11 go Step 3
12 end-if
Step 3 Interrogate with P = 1/4
13 if number_of_aircraft_replied == 0 then
14 go Step 2
15 else if number_of_aircraft_replied == 1 then
16 repeat Step 3
17 else if number_of_aircraft_replied > 1 then
18 go Step 4
19 end-if
Step 4 Interrogate with P = 1/8
20 if number_of_aircraft_replied == 0 then
21 go Step 3
22 else if number_of_aircraft_replied == 1 then
23 repeat Step 4
24 else if number_of_aircraft_replied > 1 then
25 go Step 5
26 end-if
Step 5 Interrogate with P = 1/16
27 if number_of_aircraft_replied == 0 then
28 go Step 4
29 else if (number_of_aircraft_replied == 1 or
number_of_aircraft_replied > 1) then
30 repeat Step 5
31 end-if
end procedure
In the above pseudocode, the control length is 1. This
means that when there is no reply for only one interrogation,
we increase the probability of interrogation (POI), or when
garbling happens, we decrease the POI. When control length
is n, after n consecutive interrogations with no reply, we in-
crease the POI. For probabilities 1 and 0.0625, some excep-
tions exist. Also, for probability 0.5, when detection occurs,
we increase the POI again. Typically, for detection cases,
there is no probability change.
2.1 Performance of the Algorithm
The algorithm runs with three different parameters for
each Mode-S SSR specifications, namely pulse repetition
Parameter
Selected Values
PRF (Hz)
[150, 225, 300]
RPM (1/min)
[6, 10, 15]
Beam-width (degree)
[1.2, 1.8, 2.4]
Tab. 2. Basic Mode-S SSR specifications that are used in
simulations.
frequency, RPM, and beam-width. The selected parameters
are given in Tab. 2. These parameters are selected according
to the most common values used in industrial airport radar
antennas [30].
Monte Carlo simulation with 1000 iterations is run for
each number of aircraft ranging from 2 to 20; therefore, a to-
tal of 3 × 3 × 3 ×19 = 513 combinations is investigated in
the results. Both our adaptive probabilistic interrogation al-
gorithm and theoretical static interrogations are simulated to
compare the performance in terms of time spent detecting all
aircraft in the dense air environment. The time spent collect-
ing all ICAO addresses during all-call interrogations and the
mean number of interrogations to receive aircraft replies one
by one without garbling are the most important metrics to
consider.
Using the proposed algorithm and static interrogations
in Monte Carlo simulation in order to compare the perfor-
mance of these approaches in terms of the number of all-call
interrogations, the results can be seen in Fig. 3.
The graph is divided into three different parts accord-
ing to the density of the aircraft in the target area. Interrogat-
ing with static 0.5 probability all-calls gives the best perfor-
mance in the less dense blue part since the garbling effects
are diminished due to few numbers of aircraft. Using low
probabilities such as 0.125 or 0.0625 to detect a few aircraft
in the blue part happens to be an inefficient method. Interro-
gating with static 0.25 probability gives the best result in the
dense, orange part. However, interrogation with static 0.125
or 0.0625 comes on top in the densest, green part in terms of
performance since the number of aircraft is too many and
garbling effects are immense in low probabilistic all-calls.
Considering all the static interrogation approaches, they all
tend to give good performances within defined colored parts.
Fig. 3. Overall performance of adaptive algorithm vs static
interrogations.

464 A. ONCU, A. G. AYDIN, Y. ERDOGAN, ET AL., MODE-S RADAR INTERROGATION ALGORITHM DESIGN …
However, the proposed algorithm provides the best results
when there are more than 10 aircraft in the dense environ-
ment, and moreover, it comes second best when there are
few aircraft.
2.2 Effects of Basic Radar Specifications on
Performance
Simulation results of adaptive interrogation algorithm
in terms of different RPM, PRF, and BW parameters are
shown in Fig. 4, 5, 6.
In all three figures, the time spent to detect all aircraft
in the dense garbled region increases as the number of air-
craft increases. Here, the performance metric is the time ra-
ther than the number of interrogations, and the time spent
per pulse, Tp and total time spent for complete interrogation,
Tci are calculated as follows:
p
1
TPRF
=
, (6)
ci p
T Number of completed interrogations T= ×
. (7)
Fig. 4. Effect of beam-width on performance.
Fig. 5. Effect of PRF on performance.
Fig. 6. Effect of RPM on performance.
As observed in Monte Carlo simulations with 500 iter-
ations, the number of interrogations to detect all aircraft is
not affected by parameter changes. For this reason, since the
increase in PRF will shorten the time taken for interrogation,
the total time spent will also decrease. The situation is simi-
lar in the beam-width case: As the increase in beam-width
will increase the number of interrogations per aircraft in one
tour, the number of tours taken will be shortened, which will
shorten the total elapsed time. The RPM indicates how many
seconds it takes to complete a lap. For this reason, PRF is
divided by RPS (round per second) to find the total number
of pulses sent in each round. To find the number of pulses in
each beam, the total number of pulses in one round is divided
by the total number of beams. Although the increase in RPM
causes the antenna to rotate faster, we cannot say that there
is a dominant correlation between the total interrogation
time and RPM since it reduces the number of pulses in
a beam.
For selected parameters, the total time spent is shown
in Fig. 7. The proposed algorithm performs the best after the
number of aircraft exceeds 10. It is observed that it gives
very close to the best results in other places.
Fig. 7. The performance of the algorithm and static methods
with selected parameters, i.e., PRF: 150 Hz, BW:
2.4 degrees, and RPM: 6 min-1.

RADIOENGINEERING, VOL. 31, NO. 4, DECEMBER 2022 465
2.3 Limited Lockout
The lockout condition is that when an aircraft replies to
an all-call message, the ground station locks up that aircraft
in order to interrogate with roll-call messages, and this lock-
out duration takes approximately 18 seconds [29], [31]. Dur-
ing this lockout duration, we do not interrogate the detected
aircraft once again with all-call uplink requests; therefore,
we deduce that we found an aircraft in the dense air traffic
region, and we shall not interrogate it with all-call messages.
Once the eighteen-second lockout duration ends, we interro-
gate that specific aircraft again if it reenters the radar cover-
age area. By doing this, we take into account re-entries to the
radar coverage in the vicinity of sections where two different
ground stations intersect with each other’s coverages.
In Fig. 8, the performances are given in terms of total
time spent using selected parameters. Our adaptive algo-
rithm also works as the second best most of the time in dif-
ferent density regions. Nevertheless, our adaptive algorithm
comes on top regardless since the information about the
number of aircraft in the region is not required, and this is
handy for the operator in the ground station. An additional
improvement can be made by taking care of entering and ex-
iting time instances of an aircraft in the dense region rather
than just interrogation probability switching.
In addition to the lockout condition, there is also a lock-
out override scenario in which the interrogator forces a tran-
sponder to reply to all all-calls. Stochastic interrogation
needs to be used to avoid garbling in lockout override oper-
ations. This method is called stochastic lockout override ac-
quisition (SLO) and is used when radar coverage clusters of
different ground stations with the same ICs (interrogator
code) overlap with each other [32]. While implementing
SLO, the ground station (GS) uses the different probability
of reply (PR) fields to avoid RF pollution since GS will re-
ceive an immense amount of replies due to overriding. The
PR field is 4 bits in the uplink scheme, and it can take num-
bers from 8 to 12 in binary [33].
Fig. 8. The performance of the algorithm and static methods
with selected parameters for limited lockout case, i.e.,
PRF: 150 Hz, BW: 2.4 degrees, and RPM: 6 min-1.
3. Conclusion and Discussion
The time spent or the number of interrogations sent
while gathering ICAO addresses of all aircraft in the dense
region varies for the different number of aircraft. The statis-
tical performance results of the 0.5, 0.25, 0.125, and 0.0625
probability strategies, including the proposed algorithm, in
terms of the number of all-call interrogations, can be seen in
Tab. 3.
For regions 1 (2–4 aircraft) and 2 (5–10 aircraft), the
proposed algorithm gets better as the number of aircraft in-
creases in the region, and for region 3 (11–20 aircraft), the
proposed algorithm clearly outperforms other static ap-
proaches. These results show that the proposed algorithm
can be quite useful when considering future needs and ATM
(air traffic management) or UTM (UAS traffic management)
control strategies. Such air traffic control strategies are re-
quired as the air traffic environment becomes denser year
after year. As some UAVs start to use Mode-S communica-
tion protocols as well, e.g., ads-95 ranger drones of the Swiss
Airforce, and the need for ICAO registration of this type of
vehicle increases as they enter airspaces [34]. Therefore, the
increase in air traffic reveals the importance of probabilistic
interrogations in Mode-S protocol. In this paper, the theoret-
ical and computational results of probabilistic interrogations
are obtained. At the same time, an algorithm is proposed that
both reduces the operator requirement and works well with-
out many static interrogations. The proposed algorithm
works adaptively according to the density of the aircraft.
This adaptive operating principle will facilitate the adapta-
tion to future dense air traffic environments and reduce the
detection time of the aircraft in air control and monitoring
applications.
Number
of
aircraft
0.5
method
0.25
method
0.125
method
0.0625
method Algorithm
2
4.00
6.67
12.6
24.5
4.7
3
6.67
9.03
16.0
30.6
8.3
4
10.7
11.4
19.0
35.4
11.6
5
17.1
13.9
21.8
39.6
14.9
6
27.7
16.7
24.3
43.2
18.0
7
46.0
20.0
26.9
46.6
21.2
8
78.0
23.7
29.4
49.8
24.3
9
135.0
28.1
32.0
52.8
27.6
10
237.3
33.5
34.7
55.6
30.6
11
423.5
40.0
37.4
58.4
33.7
12
764.8
47.8
40.3
61.1
37.0
13
1395
57.5
43.4
63.8
40.0
14
2565
69.6
46.7
66.4
43.0
15
4749
84.5
50.1
69.0
46.2
16
8845
103.2
53.8
71.7
49.4
17
16556
126.7
57.8
74.3
52.7
18
31119
156.2
62.1
77.0
55.6
19
58714
193.6
66.8
79.7
59.0
20
111140
240.9
71.8
82.4
62.1
Tab. 3. Number of interrogations to detect aircraft for the
constant 0.5, 0.25, 0.125, and 0.0625 probability
methods and the proposed algorithm.

466 A. ONCU, A. G. AYDIN, Y. ERDOGAN, ET AL., MODE-S RADAR INTERROGATION ALGORITHM DESIGN …
Acknowledgements
This paper is an outcome of the research work under
the research contract between the Bogazici University Tech-
nology Transfer Office and the Scientific and Technological
Research Institution of Turkey, Informatics and Information
Security Research Center.
References
[1] BAKER, J. L, ORLANDO, V.A., LINK, W.B., et al. Mode S system
design and architecture. Proceedings of the IEEE, 1989, vol. 77,
no. 11, p. 1684–1694. DOI: 10.1109/5.47731
[2] INTERNATIONAL TELECOMMUNICATION UNION.
Reception of Automatic Dependent Surveillance Broadcast via
Satellite and Compatibility Studies with Incumbent Systems in the
Frequency Band 1 087.7-1 092.3 MHz. Geneva: ITU, 2017. Report
ITU-R M.2413-0
[3] BEASLEY, B. Understanding Mode S Technology. [Online] Cited
2012-10-10 Available at AvionTEq:
https://www.avionteq.com/document/Understanding-Mode-S-
technology.pdf
[4] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. Conflict-Free Direct Routings in European
Airspace. March 1997. EEC Report No. 308.
[5] STANDFUSS, T., SCHULTZ, M. Performance assessment of
European air navigation service providers. In 2018 IEEE/AIAA 37th
Digital Avionics Systems Conference (DASC). London (UK), 2018,
p. 1–10. DOI: 10.1109/DASC.2018.8569839
[6] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. EUROCONTROL Annual Report 2016. 3 August
2017.
[7] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. EUROCONTROL Five-Year Forecast 2020-2024.
4 November 2020.
[8] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. EUROCONTROL Data Snapshot #8 on the Costs
of Air Traffic Management in Europe. 23 March 2021.
[9] CABALLERO, R. M. UAS Bulletin #2. In European Civil Aviation
Conference (ECAC). France, December 2021.
[10] HAESSIG, D. A., OGAN, R. T., OLIVE, M. “Sense and avoid” -
What’s required for aircraft safety. In IEEE Southeastcon. Norfolk
(VA, USA), 2016, p. 1–8. DOI: 10.1109/SECON.2016.7506724
[11] MINUCCI, F., VINOGRADOV, E., POLLIN, S. Avoiding
collisions at any (low) cost: ADS-B like position broadcast for
UAVs. IEEE Access, 2020, vol. 8, p. 121843–121857. DOI:
10.1109/ACCESS.2020.3007315
[12] JONÁŠ, P., JANČÍK, M., HOLODA, Š., et al. Impact of SUAS
equipped with ADS-B on 1090 MHz environment. In 2020 New
Trends in Civil Aviation (NTCA). Prague (Czech Republic), 2020,
p. 63–67. DOI: 10.23919/NTCA50409.2020.9291095
[13] BATUWANGALA, E., KISTAN, T., GARDI, A., et al.
Certification challenges for next-generation avionics and air traffic
management systems. IEEE Aerospace and Electronic Systems
Magazine, 2018, vol. 33, no. 9, p. 44–53. DOI:
10.1109/MAES.2018.160164
[14] DOOLE, M., ELLERBROEK, J., HOEKSTRA, J. M. Investigation
of merge assist policies to improve safety of drone traffic in
a constrained urban airspace. Aerospace, 2022, vol. 9, p. 1–25. DOI:
10.3390/aerospace9030120
[15] KELLERMANN, R., BIEHLE, T., FISCHER, L. Drones for parcel
and passenger transportation: A literature review. Transportation
Research Interdisciplinary Perspectives, 2020, vol. 4, p. 1–13. DOI:
10.1016/j.trip.2019.100088
[16] SHRESTHA, R., BAJRACHARYA, R., KIM, S. 6G enabled
unmanned aerial vehicle traffic management: A perspective. IEEE
Access, 2021, vol. 9, p. 91119–91136. DOI:
10.1109/ACCESS.2021.3092039
[17] YOO, L. S., LEE, J. H., KO, S. H., et al. A drone fitted with
a magnetometer detects landmines. IEEE Geoscience and Remote
Sensing Letters, 2020, vol. 17, no. 12, p. 2035–2039. DOI:
10.1109/LGRS.2019.2962062
[18] JAN, S. U., KHAN, H. U. Identity and aggregate signature-based
authentication protocol for IoD deployment military drone. IEEE
Access, 2021, vol. 9, p. 130247–130263. DOI:
10.1109/ACCESS.2021.3110804
[19] KOGA, T., MORI, K. Autonomous lockout map construction
technique for secondary surveillance radar Mode S network. In
2010 IEEE Radar Conference. Arlington (VA, USA), 2010,
p. 1439–1443. DOI: 10.1109/RADAR.2010.5494389
[20] SUN, J. The 1090 Megahertz Riddle: A Guide to Decoding Mode S
and ADS-B Signals. Delft: TU Delft OPEN Publishing, 2021. ISBN:
978-94-6366-402-8 DOI: 10.34641/mg.11
[21] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. Principles of Mode S Operation and Interrogator
Codes. 2003, March 18.
[22] MATEU, J. RADIOLOCATION: Secondary Surveillance RADAR
(SSR), Air Traffic Control Radar Beacon System(ATCRBS).
[Online] 2016. Retrieved from Universidad Politecnica de
Catalunya Barcelonatech: UPCommons:
https://upcommons.upc.edu/bitstream/handle/2117/340881/SSR.pd
f?sequence=1&isAllowed=y
[23] SUN, J., VU, H., ELLERBROEK, J., et al. pyModeS: Decoding
Mode-S surveillance data for open air transportation research. IEEE
Transactions on Intelligent Transportation Systems, 2020, vol. 21,
no. 7, p. 2777–2786. DOI: 10.1109/TITS.2019.2914770
[24] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. MODE-S Specific Services and Data Link Test
Bench. 1998, April.
[25] KOGA, T. Classification of Mode S transponders by datalink
capability. In 2014 Integrated Communications, Navigation and
Surveillance Conference (ICNS) Conference Proceedings. Herndon
(VA, USA), 2014, p. O3-1–O3-7. DOI:
10.1109/ICNSurv.2014.6820008
[26] BODART, J. Mode S Surveillance Principle. [Online] Cited 2019-
02-26. Available at EUROCONTROL:
https://www.icao.int/MID/Documents/2019/MICA/MICA-
MID%20-%20WP%2002%20-
%20Mode%20S%20Surveillance%20Principle.pdf
[27] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. EUROCONTROL Specification for European
Mode S Station (EMS). 2021, December 14.
[28] BODART, J. Radar Programming (MIP). [Online] Published 2019-
02-26. Available at EUROCONTROL:
https://www.icao.int/MID/Documents/2019/MICA/MICA-
MID%20-%20WP%2012%20-%20Radar%20programming%20-
%20MIP.pdf#search=radar%20programming%20MIP
[29] EUROPEAN ORGANIZATION FOR THE SAFETY OF AIR
NAVIGATION. Mode S transponder in an airport/A-SMGCS
environment - Clarification. [Online] Published 2005-05-03.
Available at: https://www.eurocontrol.int/publication/mode-s-
transponder-airporta-smgcs-environment-clarification
[30] PAYDAR, M. SSR Mode S Coordination Issues. [Online] Published
2010-08. Available at ICAO.INT:

RADIOENGINEERING, VOL. 31, NO. 4, DECEMBER 2022 467
https://www.icao.int/WACAF/Documents/Meetings/2011/asi_ws/
pp1_ssr_modes_coordination.pdf
[31] BRIA, O. N., GIACOMANTONE, J., VILLAGARCÍA WANZA,
H. A. Interleaving scheduling algorithm for SLM transactions in
Mode S surveillance radar. In XXIV Congreso Argentino de
Ciencias de la Computación. La Plata (Argentina), 2018, p. 1–16.
ISBN: 978-950-658-472-6 DOI: 10.1007/978-3-030-20787-8_21
[32] MAZÚCH, P., KLECUN, R. Identification of an aircraft with the
mode-S. Acta Avionica, 2013, vol. XV, no. 27, p. 1–6.
ISSN: 1335-9479
[33] KOGA, T., UEJIMA, K. Results of validation of SSR mode S
interrogator identifier code coordination. In 2009 IEEE/AIAA 28th
Digital Avionics Systems Conference. Orlando (FL, USA), 2009,
p. 4.D.6-1–4.D.6-7. DOI: 10.1109/DASC.2009.5347481
[34] SCHAFER, M., STROHMEIER, M., SMITH, M., et al. OpenSky
report 2016: Facts and figures on SSR mode S and ADS-B usage.
In 2016 IEEE/AIAA 35th Digital Avionics Systems Conference
(DASC). Sacramento (CA, USA), 2016, p. 1–9. DOI:
10.1109/DASC.2016.7778030
About the Authors …
Ahmet ÖNCÜ received a BS degree in Physics and a BS
degree in Electrical and Electronics Engineering from the
Middle East Technical University (METU), Ankara, Tur-
key, in 2001 and 2002. He received an MS degree in Micro-
wave Engineering from the Technical University of Munich,
Germany, in 2004. He received a Ph.D. degree in Frontier
Sciences from the University of Tokyo, Japan, in 2008. After
his Ph.D., he started to work as a post-doctoral researcher at
the University of Tokyo and Hiroshima University. He
joined the Department of Electrical and Electronics Engi-
neering, Bogazici University, in 2010 and is currently work-
ing as an Associate Professor. He received an outstanding
design award at the IEEE International Asian Solid-State
Circuit Conference in 2008. His research interests are circuit
designs, microwave, RF and sensor systems, computation
for electromagnetics, radar, and AI applications. He is the
founder of the Microwave Radar and Communication La-
boratory at Bogazici University.
Ahmet Günhan AYDIN received his BS degree in Electri-
cal and Electronics Engineering from Bogazici University,
Istanbul, Turkey, in 2021. He is currently pursuing his MS
degree in the same field at Bogazici University. His research
interests include radar signal processing, Mode-S, and ma-
chine learning applications for radars.
Yunus Emre ERDOĞAN was born in 1997. He is currently
a senior undergraduate student in Electrical and Electronics
Engineering, and Physics at Bogazici University. His re-
search interests include radar signal processing, GPS, data
analysis, and machine learning applications in related fields.
Artun AKDOĞAN is pursuing for a BS degree in Com-
puter Engineering from Bogazici University. His research
interests include radar signal processing, parallel processing,
cybersecurity, and computer networks.