Article

Enhancing Communication Security in Drones Using QRNG in

Frequency Hopping Spread Spectrum
J. de Curtò 1,2,3, * , I. de Zarzà 3,4 , Juan-Carlos Cano 5 and Carlos T. Calafate 5

1 Department of Computer Applications in Science & Engineering, BARCELONA Supercomputing Center,

08034 Barcelona, Spain
2 Escuela Técnica Superior de Ingeniería (ICAI), Universidad Pontificia Comillas,
28015 Madrid, Spain
3 Estudis d’Informàtica, Multimèdia i Telecomunicació, Universitat Oberta de Catalunya,
08018 Barcelona, Spain; [email protected]
4 Departamento de Informática e Ingeniería de Sistemas, Universidad de Zaragoza,
50009 Zaragoza, Spain
5 Departamento de Informática de Sistemas y Computadores, Universitat Politècnica de València,
46022 València, Spain; [email protected] (J.-C.C.); [email protected] (C.T.C.)
* Correspondence: [email protected]

Abstract: This paper presents a novel approach to enhancing the security and reliability of drone
communications through the integration of Quantum Random Number Generators (QRNG) in
Frequency Hopping Spread Spectrum (FHSS) systems. We propose a multi-drone framework that
leverages QRNG technology to generate truly random frequency hopping sequences, significantly
improving resistance against jamming and interception attempts. Our method introduces a concur-
rent access protocol for multiple drones to share a QRNG device efficiently, incorporating robust
error handling and a shared memory system for random number distribution. The implementa-
tion includes secure communication protocols, ensuring data integrity and confidentiality through
encryption and Hash-based Message Authentication Code (HMAC) verification. We demonstrate
the system’s effectiveness through comprehensive simulations and statistical analyses, including
spectral density, frequency distribution, and autocorrelation studies of the generated frequency
sequences. The results show a significant enhancement in the unpredictability and uniformity of
Citation: de Curtò, J.; de Zarzà, I.; frequency distributions compared to traditional pseudo-random number generator-based approaches.
Cano, J.-C.; Calafate, C.T. Enhancing Specifically, the frequency distributions of the drones exhibited a relatively uniform spread across
Communication Security in Drones the available spectrum, with minimal discernible patterns in the frequency sequences, indicating
Using QRNG in Frequency Hopping high unpredictability. Autocorrelation analyses revealed a sharp peak at zero lag and linear decrease
Spread Spectrum. Future Internet 2024, to zero values for other lags, confirming a general absence of periodicity or predictability in the
16, 412. https://doi.org/10.3390/
sequences, which enhances resistance to predictive attacks. Spectral analysis confirmed a relatively
fi16110412
flat power spectral density across frequencies, characteristic of truly random sequences, thereby
Academic Editor: Paolo Bellavista minimizing vulnerabilities to spectral-based jamming. Statistical tests, including Chi-squared and
Kolmogorov-Smirnov, further confirm the unpredictability of the frequency sequences generated by
Received: 8 October 2024
Revised: 31 October 2024
QRNG, supporting enhanced security measures against predictive attacks. While some short-term
Accepted: 6 November 2024 correlations were observed, suggesting areas for improvement in QRNG technology, the overall
Published: 8 November 2024 findings confirm the potential of QRNG-based FHSS systems in significantly improving the security
and reliability of drone communications. This work contributes to the growing field of quantum-
enhanced wireless communications, offering substantial advancements in security and reliability for
drone operations. The proposed system has potential applications in military, emergency response,
Copyright: © 2024 by the authors.
and secure commercial drone operations, where enhanced communication security is paramount.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
Keywords: frequency hopping spread spectrum (FHSS); quantum random number generator
distributed under the terms and
(QRNG); random processing units; quantum computing; drone communications; quantum enhanced
conditions of the Creative Commons
Attribution (CC BY) license (https://
communications
creativecommons.org/licenses/by/
4.0/).

Future Internet 2024, 16, 412. https://doi.org/10.3390/fi16110412 https://www.mdpi.com/journal/futureinternet

Future Internet 2024, 16, 412 2 of 33

1. Introduction
In an increasingly digital and interconnected world, and with the advent of sophisti-
cated AI techniques [1], the need for secure communication channels has never been more
critical. Sensitive information, whether personal, corporate, or governmental, must be
protected against a myriad of threats including eavesdropping, data interception, and delib-
erate jamming. Frequency Hopping Spread Spectrum (FHSS) [2,3] has long been employed
as a technique to enhance the security and robustness of wireless communications.
The ever-increasing demand for secure communication channels has led to the ex-
ploration of advanced techniques that can withstand interception and jamming attempts.
Frequency Hopping Spread Spectrum (FHSS) is a well-established method to enhance
communication security. This article explores the potential of using a Randomness Pro-
cessing Unit (RPU) to generate true random sequences for frequency hopping, thereby
significantly enhancing the unpredictability and robustness of communication systems. By
rapidly changing the carrier frequency according to a specific sequence, FHSS spreads the
signal over a wide spectrum, making it more resistant to interference and more difficult
to intercept.
Traditionally, the sequences used for frequency hopping are generated using pseudo-
random number generators (PRNGs). While these sequences can appear random, they are
ultimately deterministic and, given enough time and resources, can potentially be predicted
or reverse-engineered. This inherent vulnerability poses a significant risk to communication
systems, especially in environments where security is paramount. Improvements over
traditional PRNGs [4–6] have been explored in the literature, with emphasis on an efficient
implementation and portability.
The advent of Randomness Processing Units (RPUs) [7,8] offers a promising solution
to this challenge. RPUs are specialized hardware devices capable of generating true random
numbers based on inherently unpredictable physical processes, such as quantum noise or
thermal fluctuations. Unlike PRNGs, the numbers generated by RPUs are not reproducible
or predictable, even with complete knowledge of the system’s internal workings.
This manuscript explores the concept of leveraging RPUs to generate true random
sequences [9,10] for frequency hopping in communication systems. By replacing deter-
ministic sequences with truly random ones, we can significantly enhance the security
and robustness of FHSS. We will study the theoretical background of FHSS, the math-
ematical formulation of true random frequency hopping, and the security benefits and
implementation challenges associated with this approach. A preliminary use case with an
actual QRNG device setup in a multi-drone configuration is presented. This exploration
aims to highlight the importance of incorporating true randomness in securing modern
communication infrastructures.
The paper is organized as follows: Section 2 provides an overview and related works
while Section 3 provides the theoretical background and considerations for Frequency
Hopping Spread Spectrum (FHSS) and the use of Randomness Processing Units (RPUs). In
Section 4, we present our methodology for integrating an RPU to generate truly random
frequency hopping sequences, enhancing the security of FHSS systems. Section 5 details the
implementation aspects, including the setup of the RPU, generation and synchronization of
frequency hopping sequences, and security protocols. In Section 6, we apply the proposed
method to a drone cloud with a ring topology, demonstrating its benefits in multi-drone
communications. Section 7 presents the evaluations, including experimental setup, proce-
dures, and results that demonstrate the effectiveness of our approach. Section 8 explores
practical applications and potential future developments of the system. Finally, Section 9
concludes the paper and outlines future research directions.

2. Related Works
Frequency Hopping Spread Spectrum (FHSS) has long been recognized as a robust
method for secure wireless communications due to its resistance to interference and eaves-
dropping. The concept of FHSS was first patented by Lamarr and Antheil [11], where
Future Internet 2024, 16, 412 3 of 33

the carrier frequency is rapidly switched according to a specific sequence known only to
the communicating parties. Traditional FHSS systems utilize Pseudo-Random Number
Generators (PRNGs) to create these hopping sequences. While PRNGs are computation-
ally efficient, they are inherently deterministic, making them susceptible to prediction and
reverse engineering if an adversary intercepts sufficient portions of the communication [12].
The limitations of PRNGs in security-critical applications have spurred interest in True
Random Number Generators (TRNGs) and, more recently, Quantum Random Number
Generators (QRNGs). QRNGs exploit fundamental quantum mechanical processes, such
as quantum vacuum fluctuations and photon emission events, to produce truly random
numbers that are inherently unpredictable [13]. Jennewein et al. [14] demonstrated an
early implementation of a QRNG using photon polarization measurements, providing a
significant enhancement in randomness over classical TRNGs.
In the context of cryptographic systems, secure key distribution and encryption are
paramount. The RSA algorithm [15], developed by Rivest, Shamir, and Adleman, revo-
lutionized public-key cryptography by utilizing the computational difficulty of factoring
large composite numbers. RSA allows for secure key exchange over unsecured channels,
which is essential for initializing secure communications. However, RSA’s security relies
on key lengths that may be computationally intensive for resource-constrained devices.
Elliptic Curve Cryptography (ECC) [16,17] offers similar security levels to RSA but
with smaller key sizes, making it more suitable for devices with limited computational
resources, such as drones and IoT devices. ECC’s efficiency stems from the mathematical
complexity of the elliptic curve discrete logarithm problem, which remains intractable for
sufficiently large key sizes.
For symmetric encryption, the Advanced Encryption Standard (AES) [18] has become
the industry standard due to its balance of security and performance. AES is employed
for encrypting data streams after the initial key exchange, ensuring confidentiality and
integrity during communication sessions. The combination of public-key cryptography for
key exchange (RSA or ECC) and symmetric encryption (AES) provides a robust framework
for secure communications.
Recent advancements in integrating QRNGs into communication systems have shown
promise in enhancing security. Abellán et al. [19] utilized quantum phase fluctuations in a
laser diode to generate high-speed random numbers suitable for cryptographic applications.
Such QRNGs can be integrated into FHSS systems to produce truly random hopping
sequences, significantly increasing the difficulty for adversaries attempting to predict or
jam the communication.
In the field of drone communications, security is a critical concern due to the increasing
reliance on unmanned aerial vehicles (UAVs) for military, commercial, and civilian appli-
cations. Existing research has explored various aspects of secure drone communications,
including secure routing protocols [20], intrusion detection systems [21], and encryption
schemes tailored for UAV networks [22]. However, the application of QRNG-based FHSS
in drone networks remains relatively unexplored.
Frequency hopping in wireless communication systems, while providing security
and robustness against interference and eavesdropping, introduces challenges related to
scheduling and potential interruption of connections. Efficient scheduling is crucial to
ensure seamless communication, especially in dynamic networks such as multi-hop UAV
systems [23] and IoT Networks [24].
Our work builds upon these foundational studies by integrating QRNGs into FHSS
systems specifically for multi-drone networks. By employing secure key exchange mecha-
nisms such as RSA or ECC for the initial distribution of synchronization information and
AES for encrypting the frequency hopping sequences, we address the challenges associ-
ated with synchronization and security in true random FHSS systems. This integration
enhances the unpredictability of the hopping sequences and provides a robust defense
against jamming and interception attempts.
Future Internet 2024, 16, 412 4 of 33

Furthermore, the implementation of a concurrent access protocol and a shared memory

system for random number distribution ensures that multiple drones can efficiently share
a single QRNG device without compromising the quality of randomness. This approach
optimizes resource utilization in scenarios where equipping each drone with its own QRNG
may be impractical due to size, weight, or power constraints.
By combining quantum randomness with established cryptographic practices, our
methodology offers a significant advancement in securing drone communications. It
addresses the vulnerabilities associated with deterministic PRNGs and enhances the overall
resilience of the communication network against sophisticated attacks.

3. Theoretical Background and Considerations

In secure communication systems, protecting the integrity and confidentiality of trans-
mitted data is crucial. FHSS is a technique used to achieve these goals by distributing
the signal across a wider bandwidth, thereby making it more resistant to interference and
eavesdropping. This section provides a detailed mathematical foundation for FHSS, ex-
plores the generation of frequency hopping sequences, and introduces the concept of using
Randomness Processing Units (RPUs) to enhance security through true random sequences.

3.1. Frequency Hopping Spread Spectrum (FHSS)

FHSS operates by changing the carrier frequency of a signal at regular intervals
according to a specific hopping sequence. The signal S(t) is transmitted across different
frequencies over time, making it difficult for an unauthorized party to intercept or jam the
communication. The mathematical model of the transmitted signal can be expressed as:
N
S(t) = ∑ Ao cos(2π f o t + ϕo ) · 1[to ,to+1 ) (t) (1)
o =1

where:
• Ao is the amplitude of the signal during the o-th hop,
• f o is the carrier frequency for the o-th time interval,
• ϕo is the phase for the o-th time interval,
• 1[to ,to+1 ) (t) is an indicator function that is 1 if t ∈ [to , to+1 ) and 0 otherwise,
• to and to+1 are the start and end times of the o-th hop, respectively,
• N is the total number of frequency hops.

Frequency Hopping Sequence

The sequence of frequencies { f o } used in FHSS is typically generated using a PRNG.
Let {ro } be a sequence of pseudo-random numbers generated by a PRNG. The frequency
for the o-th hop can be determined as:
ro
f o = f min + ( f max − f min ) (2)
R
where:
• f min and f max are the minimum and maximum frequencies in the hopping band,
• ro is the pseudo-random number for the o-th hop, and
• R is the range of the PRNG (i.e., the maximum possible value of ro ).
The key challenge with PRNG-based sequences is that, given enough output sam-
ples, an adversary can potentially reconstruct the sequence generation process, posing a
security risk.

3.2. True Random Frequency Hopping with RPUs

RPUs are hardware devices designed to produce true random numbers by harnessing
unpredictable physical phenomena, such as quantum noise or thermal fluctuations. These
Future Internet 2024, 16, 412 5 of 33

numbers are fundamentally unpredictable and cannot be reproduced, providing a higher

level of security compared to pseudo-random sequences.
Let { Ro } denote a sequence of true random numbers generated by an RPU. The
corresponding frequency hopping sequence can be represented as:

Ro
f o = f min + ( f max − f min ) (3)
Rmax
where:
• Ro is the true random number for the o-th hop,
• Rmax is the maximum value that Ro can take.
The entropy H of the true random sequence { Ro } is given by:

H = − ∑ P( Ro ) log P( Ro ) (4)
o
1
In the case of uniformly distributed random numbers, P( Ro ) = Rmax for all o, lead-
ing to:

H = log Rmax (5)

This maximum entropy ensures that the sequence is completely unpredictable.
The process of generating and synchronizing the frequency hopping sequence is
depicted in Figure 1. This flowchart indicates the steps from generating true random
numbers to initiating FHSS communication.

Start

Generate True Random

Numbers using QRNG

Map Random Numbers

to Frequencies

Synchronize
Transmitter and Receiver

Begin FHSS
Communication

End

Figure 1. Flowchart of the frequency hopping sequence generation and synchronization process.

The primary security advantage of using RPUs lies in the true randomness of the
generated sequences. Even if an adversary has access to the transmitted signal, they
cannot predict future frequency hops without direct access to the RPU or the shared
random sequence. This feature significantly enhances the robustness of the system against
interception and jamming.

3.3. Statistical Properties and Analysis

To evaluate the statistical properties of the sequence, we can analyze the autocorrela-
tion function, which measures the similarity between the sequence at different time shifts.
For a sequence { f o }, the autocorrelation function R f (τ ) is defined as:

R f (τ ) = E[ f o f o+τ ] − E[ f o ]E[ f o+τ ] (6)

Future Internet 2024, 16, 412 6 of 33

For a true random sequence, R f (τ ) ≈ 0 for τ ̸= 0, indicating that there is no predictable

pattern over time.
The power spectral density (PSD) of the frequency-hopped signal provides insight
into how the signal’s power is distributed across the frequency spectrum. The PSD of a
signal S(t) is defined as:
2
"Z #
1 T/2
− j2π f t
SSS ( f ) = lim E S(t)e dt (7)
T →∞ T − T/2

For a system using true random FHSS, the PSD is expected to be flat across the hopping
band, indicating uniform distribution of signal power.
This detailed mathematical foundation underscores the robustness and security advan-
tages of using true random sequences for frequency hopping, as opposed to deterministic
or pseudo-random sequences. The next sections will delve into the security analysis and
practical considerations for implementing such a system.

4. Methodology and System Architecture

In this section, we present our methodology to enhancing the security and robustness
of FHSS systems by integrating a Randomness Processing Unit (RPU) to generate truly
random frequency hopping sequences. The core idea of our methodology is to use the RPU
to produce a sequence of true random numbers that are used to determine the hopping
frequencies. Both the transmitter and receiver have access to this same sequence of random
numbers, enabling them to stay synchronized during communication. This is achieved
through secure key distribution or pre-sharing the random sequence prior to communi-
cation. By using true random numbers, the frequency hopping pattern becomes entirely
unpredictable to an adversary, significantly enhancing security against eavesdropping and
jamming attacks.
Unlike traditional FHSS systems that rely on pseudo-random number generators
(PRNGs) and shared seeds, which can be potentially predicted or replicated by attack-
ers, our system’s use of true randomness ensures that the hopping sequence cannot be
reproduced without direct knowledge of the random numbers generated by the RPU. The
synchronization between the transmitter and receiver does not depend on generating the
same sequence independently but on both parties accessing the same random sequence
securely. Secure key distribution mechanisms, such as public-key cryptography or pre-
shared keys, are employed to share the random sequence or the necessary keys to access
it. This approach maintains synchronization and ensures that only authorized parties can
participate in the communication, thereby enhancing overall system security.
Figure 2 illustrates the overall system architecture, highlighting the key components
and their interactions within the proposed method.

(1) Randomness
(2) True Random
Processing Unit (3) FHSS Modulator (4) Transmitter
Frequency Sequence
(RPU)

Wireless Channel

(7) Secure Key (6) Synchronization

(5) Receiver
Distribution Mechanism

Figure 2. Diagram illustrating the methodology of using a Randomness Processing Unit (RPU) for
true random frequency hopping in an FHSS system.

In FHSS, the carrier frequency is changed according to a predefined sequence [25,26],

hopping between different frequencies in a seemingly random fashion. This method
spreads the signal over a wider bandwidth, providing robustness against narrowband
interference and eavesdropping. The sequence of frequency hops is traditionally generated
Future Internet 2024, 16, 412 7 of 33

using pseudo-random number generators (PRNGs), which, while complex, are ultimately
deterministic and potentially predictable [27,28].
Next, we provide a detailed explanation of how each component interacts and the
data flow between them.
1. Randomness Processing Unit (RPU): The RPU generates a stream of true random
numbers by exploiting inherent physical phenomena such as quantum noise or ther-
mal fluctuations. These random numbers form the basis of the frequency hopping
sequence and are critical for ensuring unpredictability in the system.
2. True Random Frequency Sequence Generator: The random numbers produced by the
RPU are processed to generate a sequence of frequencies { f o }. This involves mapping
the random numbers to specific frequencies within the available frequency band F.
The mapping ensures a uniform probability distribution over F, maximizing entropy.
The frequency sequence generator interfaces directly with the RPU to receive the
random numbers and with the FHSS modulator to provide the frequency sequence.
3. FHSS Modulator: The FHSS modulator receives the true random frequency sequence
from the frequency sequence generator. It modulates the carrier signal according
to the frequency f o specified for each time interval [to , to+1 ). The modulator also
incorporates the data to be transmitted, applying appropriate modulation techniques
(e.g., frequency shift keying or phase shift keying). The modulated signal is then
passed to the transmitter.
4. Transmitter: The transmitter amplifies the modulated signal and broadcasts it over
the wireless channel. It ensures that the signal conforms to regulatory standards for
transmission power and spectral occupancy. The transmitter interfaces with the FHSS
modulator for input and communicates over the wireless channel to the receiver.
5. Receiver: On the receiving end, the receiver captures the signal from the wireless
channel. It must be synchronized with the transmitter to correctly follow the frequency
hopping sequence. The receiver demodulates the signal using the same frequency
sequence { f o }, extracting the transmitted data. The receiver interfaces with the
synchronization mechanism to maintain alignment with the transmitter’s frequency
hopping pattern.
6. Synchronization Mechanism: The synchronization mechanism is critical for main-
taining alignment between the transmitter and receiver. It ensures that both ends
use the same frequency f o at any given time t. Synchronization information, such
as initial timing parameters and any required updates, is communicated through a
secure channel or embedded within the communication protocol. The synchronization
mechanism interfaces with both the frequency sequence generator (to obtain or verify
the frequency sequence) and the receiver (to adjust timing and sequence alignment).
7. Secure Key Distribution: The secure key distribution component facilitates the ex-
change of keys or synchronization information necessary for secure communication. It
interfaces with the synchronization mechanism to provide the required cryptographic
keys or parameters. Secure channels, possibly using public-key cryptography, are
employed to prevent interception or tampering during key exchange.
The data flow begins with the RPU generating true random numbers, which are
converted into a frequency sequence by the frequency sequence generator. This sequence
is fed into the FHSS modulator, along with the data to be transmitted. The FHSS modu-
lator outputs a frequency-hopped, modulated signal to the transmitter. The transmitter
broadcasts the signal over the wireless channel to the receiver. The receiver, synchronized
via the synchronization mechanism, captures and demodulates the signal using the same
frequency sequence to retrieve the data. Throughout this process, the synchronization
mechanism and secure key distribution ensure that both ends maintain alignment and that
the communication remains secure. Implementing dynamic control architectures and ro-
bust communication protocols is essential for swarm unmanned aerial vehicles to function
effectively [29].
Future Internet 2024, 16, 412 8 of 33

Let S(t) represent the transmitted signal. In a traditional FHSS system, the carrier
frequency at time t is determined by a sequence { f o }, where o = 1, 2, . . . , N and N is the
total number of hops in the sequence. The frequency-hopped signal can be expressed as:

N
S(t) = ∑ Ao cos(2π f o t + ϕo ) · 1[to ,to+1 ) (t)
o =1
.
Here, Ao represents the amplitude, ϕo the phase, and 1[to ,to+1 ) (t) is an indicator func-
tion defining the time interval for each hop.

4.1. True Random Frequency Hopping

With an RPU, the sequence { f o } is generated by a truly random process. If f o is chosen
uniformly from a set of available frequencies F, the probability distribution of selecting any
particular frequency is:
1
P( f o = f z ) = (8)
| F|
where | F | is the cardinality of the set F. The unpredictability of the frequency hops,
characterized by entropy H, is maximized:

H=− ∑ P( f ) log P( f ) = log | F| (9)

f ∈F

This high entropy ensures that even with complete knowledge of the hopping algo-
rithm, an adversary cannot predict the sequence, as it is truly random.
The primary security advantage of true random frequency hopping lies in its anti-
jamming capabilities. Given that the sequence is entirely unpredictable, a jammer must
either cover the entire bandwidth (which requires considerable power) or fail to disrupt the
communication. The jamming margin, defined as the difference between the bandwidth of
the spread spectrum and the jamming bandwidth, is maximized.
The security of the system is further enhanced because eavesdroppers cannot anticipate
the next frequency hop. The information-theoretic security provided by the truly random
sequence means that even with infinite computational power, an adversary cannot predict
future hops without the actual random sequence.
The primary challenge with true random FHSS is synchronization between the trans-
mitter and receiver. Both ends must be perfectly synchronized in terms of the hopping
sequence. This can be achieved through secure key distribution mechanisms or synchro-
nization protocols.
The true random sequence or the seed value for the RPU must be securely shared
between the communicating parties. This can be done using public-key cryptography or
other secure channels.
The implementation of our RPU-based FHSS system presents both challenges and
opportunities for enhancing communication security. One key consideration is the rate
at which the RPU can generate random numbers, as this directly impacts the frequency
hopping rate and, consequently, the system’s resilience to jamming attempts. To address
potential latency issues, we employ a buffer system that pre-generates a pool of random
frequencies, ensuring a continuous supply for rapid hopping. Additionally, we implement
a dynamic hopping rate that adapts to the perceived threat level; under normal conditions,
the system may use a moderate hopping rate to conserve power, but it can switch to a more
aggressive rate when jamming or interception is suspected. The RPU’s true randomness
also allows for the implementation of a frequency exclusion list, dynamically avoiding
frequencies that are detected to be under interference, without compromising the unpre-
dictability of the overall sequence. This adaptive approach, coupled with the inherent
unpredictability of the RPU, provides a significant advantage over traditional PRNG-based
systems in terms of security and flexibility.
Future Internet 2024, 16, 412 9 of 33

4.2. Buffer System Implementation

To address potential latency issues associated with real-time random number gen-
eration, we implement a buffer system that pre-generates and stores a pool of random
frequencies. The buffer operates as a circular queue with a size B, determined based on
the maximum expected hopping rate and the RPU’s generation capabilities. The algorithm
for buffer management, as described in Algorithm 1, ensures that the buffer is replenished
proactively, maintaining a threshold level Bmin to prevent underrun. This approach de-
couples the random number generation process from the frequency hopping mechanism,
allowing for continuous operation even if momentary delays occur in the RPU output.

Algorithm 1 Algorithm for Buffer Management

1: Initialization:
2: Set buffer size B and minimum threshold Bmin .
3: Initialize buffer pointer p ← 0.
4: while True do
5: if Buffer_Size ≤ Bmin then ▷ Buffer Replenishment
6: Generate random number Ro from RPU.
7: Map Ro to frequency f o .
8: Store f o in buffer at position p.
9: Update pointer p ← ( p + 1) mod B.
10: end if
11: if Time to Hop then ▷ Frequency Hopping
12: Retrieve next frequency f hop from buffer.
13: Transmit using frequency f hop .
14: end if
15: end while

4.3. Dynamic Hopping Rate Implementation

The system can adjust the frequency hopping rate Rh based on the perceived threat
level. A threat detection module monitors indicators such as increased error rates, signal
interference patterns, and unexpected signal strength variations. If the module detects
anomalies exceeding predefined thresholds, it can signal the control unit to increase Rh
within allowable limits [ Rhmin , Rhmax ]. Conversely, in low-threat environments, Rh is de-
creased to conserve energy.

Threat Detection Algorithm

The threat detection algorithm operates as described in Algorithm 2.

Algorithm 2 Threat Detection and Hopping Rate Adjustment

1: while Communication is Active do
2: Monitor communication metrics:
• Bit Error Rate (BER)
• Signal-to-Noise Ratio (SNR)
• Received Signal Strength Indicator (RSSI)
3: if Any metric exceeds threshold then
4: Set threat level Threat_Level ← High.
5: Adjust hopping rate Rh ← min( Rh + ∆R, Rhmax ).
6: else
7: Set threat level Threat_Level ← Low.
8: Adjust hopping rate Rh ← max( Rh − ∆R, Rhmin ).
9: end if
10: end while
Future Internet 2024, 16, 412 10 of 33

To clarify the operation of the Threat Detection Algorithm and address concerns about
how often the system updates its threat assessment and adjusts the frequency hopping rate,
we provide additional details on the timing and responsiveness of the mechanism.
The Threat Detection Algorithm can operate continuously during active communica-
tion, monitoring key communication metrics such as Bit Error Rate (BER), Signal-to-Noise
Ratio (SNR), and Received Signal Strength Indicator (RSSI). The assessment of these metrics
occurs at regular intervals defined by the system’s monitoring frequency f monitor , which
can be configured based on the specific requirements and capabilities of the system. In
high-security environments, f monitor can be set to evaluate metrics every hop interval,
allowing the system to detect threats and respond in real-time.
Upon detecting that any monitored metric exceeds its predefined threshold, indicating
a potential attack or interference, the system increases the frequency hopping rate Rh by
a step size ∆R. The adjustment is immediate and can occur within a single hop interval,
enabling the system to respond swiftly to threats. The time it takes to adjust from the
minimum hopping rate Rhmin to the maximum hopping rate Rhmax depends on the value
of ∆R and the frequency of metric assessments. For example, if ∆R is set such that Rh
increases by a fixed percentage or value each time a threat is detected, and assessments
occur every hop, the total time Tadjust to reach Rhmax is:

Rhmax − Rhcurrent
 
Tadjust = × Thop (10)
∆R
where Rhcurrent is the current hopping rate at the time of the first threat detection, and
Thop = R1 is the duration of each hop.
h
To prevent rapid oscillations of the hopping rate due to transient fluctuations or
false positives, the system may incorporate hysteresis or require that the thresholds be
exceeded for a certain number of consecutive assessments before adjusting Rh . Additionally,
when the threat subsides—indicated by metrics returning below threshold levels—the
system decreases Rh gradually to conserve resources, ensuring a balance between security
and efficiency.
In our current implementation, the dynamic adjustment of Rh in response to detected
threats is conceptual and serves to illustrate how the methodology could be applied in
real-world scenarios. We consider the integration of real-time threat detection and hopping
rate adjustment to be feasible with appropriate hardware and software optimizations.
Implementing this functionality would involve configuring the monitoring frequency
f monitor , defining appropriate thresholds for the metrics, and determining suitable values
for ∆R based on the system’s performance requirements and operational environment.

4.4. Frequency Exclusion List Management

Interference detection can be performed using real-time spectrum analysis. Frequen-
cies exhibiting high noise levels or consistent transmission errors are added to a frequency
exclusion list Fexcl . The system updates the set of available frequencies Favail = F \ Fexcl
dynamically, where the operator “\” denotes the set difference between two sets. This
means that Favail consists of all frequencies in F that are not in Fexcl . By removing the
excluded frequencies from the full set, we ensure that only reliable frequencies are used for
communication. To maintain randomness, the RPU can be informed of Favail so that it only
maps random numbers to permissible frequencies without altering the underlying entropy
of the sequence.

Interference Detection and Frequency Update Algorithm

To ensure that randomness is preserved across the adjusted set of available frequencies,
the mapping function within the Random Processing Unit (RPU) is modified to account
for the updated set of frequencies, Favail , as described in Algorithm 3. This adjustment
process involves normalizing the generated random numbers to match the range of Favail ,
thus maintaining a uniform probability distribution over the entire frequency set. The
Future Internet 2024, 16, 412 11 of 33

mapping is performed by selecting a frequency f o from Favail in a manner that corresponds

to each random number Ro , ensuring uniform randomness across the set. This process can
be described as follows:
 
Ro
f o = Favail × | Favail | (11)
Rmax
where:
• Ro is the raw random number generated by the RPU,
• Rmax is the maximum possible value of Ro ,
• | Favail | is the cardinality (size) of the available frequency set Favail ,
• and f o is the selected frequency within Favail .
This approach scales each random number Ro to a valid index within Favail , effectively
normalizing the random output for uniform selection across the updated frequency set.
This normalization ensures the distribution of frequency selections remains unbiased and
uniformly spread over Favail .

Algorithm 3 Interference Detection and Frequency Exclusion

1: while Communication is Active do
2: for all Frequencies f ∈ F do
3: Measure interference level I ( f )
4: if I ( f ) > Ithreshold then
5: Add f to Fexcl
6: else
7: Remove f from Fexcl if present
8: end if
9: end for
10: Update available frequencies Favail ← F \ Fexcl
11: Inform RPU of updated Favail
12: end while

4.5. Enhanced Security Analysis

The integration of true random frequency hopping introduces several security benefits.
We analyze potential attack scenarios and discuss mitigation strategies.

4.5.1. Attack Scenarios and Mitigations

• Brute-Force Attacks: An adversary attempts to scan all possible frequencies to intercept
or jam the communication.
Mitigation: The large frequency band and rapid hopping rate make this impractical
due to the required resources and time.
• Replay Attacks: Captured signals are replayed to disrupt communication or trick the
receiver into accepting false data.
Mitigation: Implementing time-stamped authentication tokens prevents the accep-
tance of outdated or replayed signals.
• Synchronization Attacks: In the context of our system, synchronization between the
transmitter and receiver is crucial due to the use of true random frequency hopping
sequences. An adversary may attempt to desynchronize the communication by intro-
ducing delays, jamming synchronization signals, or injecting false synchronization
messages. Desynchronization can lead to loss of communication and reduced sys-
tem reliability.
Mitigation: To counter synchronization attacks, we could employ robust synchroniza-
tion protocols that include the following features:
– Error Correction Codes: Utilizing error correction codes (ECC) in synchroniza-
tion messages allows the receiver to detect and correct errors introduced by
Future Internet 2024, 16, 412 12 of 33

interference or malicious activities, ensuring that synchronization data is accu-

rately received.
– Secure Authentication: Synchronization messages can be secured using crypto-
graphic techniques such as digital signatures or message authentication codes
(MACs). This prevents attackers from injecting false synchronization signals or
modifying legitimate ones.
– Periodic Resynchronization: The system could perform periodic resynchroniza-
tion at predefined intervals or upon detecting anomalies in communication. This
helps to realign the transmitter and receiver in case of minor desynchronization
and reduces the window of opportunity for attackers.
– Redundant Synchronization Channels: Employing multiple synchronization
channels or methods (e.g., combining time-based synchronization with signal-
based cues) could add redundancy, making it more difficult for an attacker to
disrupt all synchronization paths simultaneously.
– Monitoring and Anomaly Detection: Continuously monitoring communication
metrics (e.g., synchronization error rates, unexpected delays) could allow the sys-
tem to detect potential synchronization attacks early and initiate countermeasures,
such as increasing the hopping rate or switching to a secure fallback mode.

4.5.2. Information-Theoretic Security

The use of true random sequences ensures that the mutual information between the
transmitted signal and any intercepted signal by an adversary is minimized. The entropy
H of the frequency sequence is maximized, making the system secure against attackers
with unlimited computational resources.

I (Transmitted Signal; Intercepted Signal) = 0 (12)

This implies that without knowledge of the true random sequence or successful
synchronization, an adversary gains no useful information from intercepted signals.

4.5.3. Latency and Throughput Models

In communication systems, latency L refers to the time delay experienced by data as it
travels through the system. It can be expressed as:

L = Lprop + Lproc + Lsync + Lqueue (13)

where:
• Lprop is the propagation delay,
• Lproc is the processing delay,
• Lsync is the synchronization delay,
• Lqueue is the queuing delay.
While latency is an important parameter affecting the system’s responsiveness, through-
put T is a measure of how much data can be successfully transmitted over the communi-
cation channel per unit time. It is influenced by factors such as bandwidth, modulation
schemes, coding rates, and error rates.
The throughput can be modeled as:

T = R × (1 − Pe ) (14)
where:
• R is the raw data rate (bits per second),
• Pe is the bit error rate (BER).
The raw data rate R depends on the bandwidth B and the spectral efficiency η (mea-
sured in bits per second per Hertz):
Future Internet 2024, 16, 412 13 of 33

R = B×η (15)
In the context of our FHSS system with true random frequency hopping, the hopping
rate Rh can impact the effective data rate due to factors such as the time required for
frequency switching and synchronization overhead. The effective data rate Reff can be
adjusted to account for these factors:
!
tswitch + tsync
Reff = R × 1 − (16)
Thop

where:
• tswitch is the time taken to switch frequencies,
• tsync is the time spent on synchronization during each hop,
• Thop is the duration of each hop (i.e., Thop = R1 ).
h
The throughput then becomes:

T = Reff × (1 − Pe ) (17)
This formulation acknowledges that while latency affects the timing of data delivery,
throughput is primarily determined by the data rate and the error performance of the
system. High latency does not necessarily reduce the throughput but can impact the overall
system performance in terms of delay-sensitive applications.
It is important to understand that in practical FHSS systems, certain physical and
protocol-level factors can introduce delays during frequency hopping.
Firstly, although the transmitter and receiver are synchronized in terms of timing
and frequency hopping sequence, the actual process of switching frequencies involves
hardware components that require a finite amount of time to change from one frequency
to another. This time is known as the frequency switching time tswitch , and it can vary
depending on the specific hardware used in the system. During this switching period,
the transmitter and receiver may not be able to send or receive data, resulting in a brief
interruption in communication.
Secondly, maintaining synchronization often involves the exchange of synchronization
signals or the execution of synchronization algorithms, which consume processing time
tsync . Even with precise synchronization, periodic adjustments and computations are
necessary to compensate for clock drift and other factors, introducing processing delays.
Therefore, the total effective time available for data transmission during each hop is
reduced by the sum of tswitch and tsync . This reduction in available transmission time per
hop effectively decreases the data rate, which is why we adjust the effective data rate Reff
to account for these factors in the throughput model.
Including tswitch and tsync in the calculations allows for a more accurate representation
of the system’s performance, especially when considering high hopping rates Rh . As Rh
increases (i.e., the duration of each hop Thop decreases), the proportion of time spent on
frequency switching and synchronization relative to the total hop duration becomes more
significant, potentially impacting the overall throughput.
In practical implementations, the latency and throughput models are primarily utilized
during the design, configuration, and optimization phases of the FHSS system. These
models help in predicting system performance under various operating conditions and in
making informed decisions about parameter settings such as hopping rate Rh , bandwidth
B, and modulation schemes.
During active communication, the system may not need to run these calculations
continuously. Instead, they can be computed periodically or when there is a significant
change in system parameters or operating conditions. For example, if the system detects
a change in the bit error rate Pe due to interference or other environmental factors, it
may recalculate the effective data rate Reff and throughput T to assess the impact on
Future Internet 2024, 16, 412 14 of 33

performance. This recalculation can trigger adjustments to system parameters to maintain

optimal operation.
The frequency of running these calculations depends on the specific requirements
and constraints of the system. In scenarios where the communication environment is
stable, the models may be recalculated infrequently or only when manual adjustments are
made. In contrast, in dynamic or high-security environments where conditions can change
rapidly, the system might perform these calculations more frequently—potentially after
every significant event or at regular intervals defined by a monitoring frequency f calc .
It is important to note that the calculations themselves are not computationally inten-
sive and can be executed quickly without imposing significant overhead on the system.
Modern communication systems often include real-time monitoring and control software
capable of performing these calculations efficiently. By balancing the frequency of these
calculations with system performance requirements, the FHSS system can maintain high
data throughput and low latency while adapting to changing conditions.
In our current setup, we did not implement the models for latency in the operational
system. The focus of our implementation was on demonstrating the feasibility of integrating
true random number generation into the FHSS mechanism and analyzing its impact on
security and randomness properties. The latency and throughput considerations are
included for comprehensiveness and to guide future work, where these factors could be
empirically measured and optimized in real-world scenarios.

4.5.4. Error Rate Analysis

The bit error rate (BER) Pb in a frequency hopping system is given by:
s !
2Eb
Pb = Q (18)
N0

where Eb is the energy per bit, N0 is the noise power spectral density, and Q(·) is the
Q-function.
Frequency hopping spreads the signal over multiple frequencies, effectively reducing
N0 and improving Pb .

5. Implementation
To implement true random frequency hopping in communication systems, several
critical components and processes must be established. This section describes the practical
implementation, including the setup of the Randomness Processing Unit (RPU), the gen-
eration and synchronization of frequency hopping sequences, and the security protocols
necessary to maintain the integrity and confidentiality of the communication.

5.1. Setup of the Randomness Processing Unit (RPU)

The RPU is a specialized hardware device that generates true random numbers. It typ-
ically leverages physical phenomena such as quantum noise, thermal noise, or radioactive
decay, which are inherently unpredictable. The RPU’s design includes:
• Noise Source: The physical process generating randomness.
• Entropy Extraction: Digital circuitry or algorithms to convert raw physical signals into
uniformly distributed random numbers.
• Post-Processing: Additional steps to ensure the uniformity and independence of the
generated numbers, such as whitening algorithms or hash functions.
Before deployment, the RPU must be rigorously tested to ensure the quality and true
randomness of the output. Statistical tests, such as the National Institute of Standards and
Technology (NIST) suite, are used to assess randomness. Key metrics include:
• Uniformity: Distribution of numbers should be uniform across the possible range.
• Independence: No correlation should exist between consecutive numbers.
• Entropy: High entropy indicates that the output is as unpredictable as possible.
Future Internet 2024, 16, 412 15 of 33

5.2. Generation and Synchronization of Frequency Hopping Sequences

Using the RPU, true random numbers Ro are generated and mapped to frequency
values f o within a predefined band. The frequency for the o-th time interval is given by:

Ro
f o = f min + ( f max − f min )
Rmax
where Rmax is the maximum possible output of the RPU. This mapping ensures that the
frequencies are chosen randomly and uniformly over the band.
A critical challenge in FHSS systems, especially when using true random frequency
hopping sequences, is maintaining synchronization between the transmitter and receiver.
Unlike traditional FHSS systems that use pseudo-random number generators (PRNGs)
with shared seeds, our system relies on sequences that are inherently unpredictable and
cannot be reproduced by generating devices independently. Therefore, both the transmitter
and receiver must have access to the same sequence of random numbers or be able to
generate them in a synchronized manner.
To achieve secure and reliable synchronization in our system, we implement a two-
stage process involving public-key cryptography and symmetric key encryption:
• Initial Synchronization and Key Agreement: At the start of communication, the
transmitter and receiver engage in a secure key exchange protocol using public-key
cryptography, such as RSA or Elliptic Curve Cryptography (ECC). This allows them
to securely agree upon a symmetric key without exposing it to potential eavesdroppers.
The symmetric key established during this phase is used for encrypting subsequent
communications, including the true random sequence and synchronization information.
• Secure Distribution of Random Sequences: With the symmetric key in place, the
transmitter securely transmits the necessary synchronization information and the
initial segment of the true random frequency hopping sequence to the receiver. This
information is encrypted using the symmetric key, ensuring that only authorized
parties with the corresponding key can decrypt and access the sequence. The use of a
symmetric key for encryption at this stage is crucial because symmetric algorithms like
Advanced Encryption Standard (AES) are more efficient for handling larger amounts
of data in real-time communication compared to public-key algorithms.
Once the initial synchronization is achieved, both parties use the shared true random se-
quence for frequency hopping. To maintain synchronization over time, the system incorporates:
• Periodic Resynchronization: Due to potential clock drift or communication delays,
periodic resynchronization is necessary. The transmitter periodically sends encrypted
synchronization beacons or timestamps using the established symmetric key. This
allows the receiver to adjust its timing and ensures continuous alignment with the
transmitter’s hopping sequence.
• Dynamic Key Updating: To enhance security, the symmetric key used for encryp-
tion can be updated periodically using key derivation functions or additional key
exchange protocols. This reduces the risk of key compromise over extended periods
and mitigates potential replay or interception attacks.
• Robust Error Handling: The synchronization protocol includes mechanisms for error
detection and correction to handle data loss or corruption due to channel impairments
or malicious interference. This ensures that synchronization integrity is maintained
even in adverse conditions.
Future Internet 2024, 16, 412 16 of 33

That is, to achieve secure and reliable synchronization in our system, we implement
a two-stage process involving public-key cryptography and symmetric key encryption.
Initially, the transmitter and receiver engage in a secure key exchange protocol using
public-key cryptography to establish a symmetric key without exposing it to potential
eavesdroppers. With the symmetric key in place, the transmitter securely transmits the nec-
essary synchronization information and the initial segment of the true random frequency
hopping sequence to the receiver using symmetric encryption. Once the initial synchroniza-
tion is achieved, both parties use the shared true random sequence for frequency hopping.
To maintain synchronization over time, the system incorporates periodic resynchroniza-
tion, dynamic key updating, and robust error handling mechanisms, ensuring continuous
alignment and enhanced security.
By integrating public-key cryptography for initial key agreement and symmetric
key encryption for ongoing secure communication, our system effectively combines the
strengths of both cryptographic approaches. Public-key cryptography facilitates secure key
exchange without prior shared secrets, while symmetric key encryption offers efficiency for
real-time data protection.
It is important to note that while the use of public-key cryptography to establish a
symmetric key is a standard practice, the unique aspect of our system lies in the secure
distribution and synchronization of true random frequency hopping sequences. Unlike
PRNG-based systems where the receiver can independently generate the same sequence us-
ing a shared seed, our approach requires careful management of the true random sequences.
The encrypted transmission of these sequences ensures that they remain confidential and
that only authorized receivers who possess the correct symmetric key can participate in the
communication. This adds a layer of security, making it significantly more challenging for
adversaries to predict or intercept the frequency hopping pattern.

5.3. Implementation Considerations and Optimization

The choice of the frequency band and the allocation strategy depends on regulatory
requirements and the specific application. The system must ensure minimal interference
with other communications, and comply with local spectrum regulations.
The efficiency of the system is influenced by the power requirements of the RPU and
the overall system. Optimizing the power usage while maintaining high security is crucial,
especially in portable or battery-operated devices.
Real-time generation and processing of random numbers and frequency hopping
require efficient algorithms and hardware. The system must ensure low latency to avoid
delays in communication.

5.4. Security and Performance Evaluation

The system’s security is evaluated based on its resistance to various types of attacks:
• Eavesdropping: The unpredictability of true random sequences makes it extremely
difficult for unauthorized parties to intercept and understand the communication.
• Jamming: True random frequency hopping enhances resistance to jamming, as the
attacker cannot predict the frequency changes.
Performance is assessed in terms of:
• Throughput: The rate at which data can be successfully transmitted.
• Latency: The delay introduced by frequency hopping and synchronization processes.
• Reliability: The system’s ability to maintain communication integrity under different
conditions, including adversarial attacks.
Future Internet 2024, 16, 412 17 of 33

6. Application to a Drone Cloud with Ring Topology

In the context of a drone cloud, a group of drones intercommunicates using a network
configuration known as a ring topology, where each drone communicates directly with
its two neighbors. This setup is particularly advantageous for coordinating movements,
sharing sensor data, and maintaining a cohesive network structure. The application of true
random frequency hopping in this environment can significantly enhance the security and
reliability of communication, especially in scenarios where the drone cloud operates in
hostile or adversarial conditions.
The communication infrastructure of the drone cloud is represented in Figure 3,
showing how drones are interconnected in a ring topology and utilize FHSS with true
random frequency hopping. The ring topology has been shown to improve communication
efficiency and fault tolerance in networked systems [30], which is particularly beneficial for
our drone cloud configuration.

FHSS Link
Drone 1 Drone 2

FHSS Link FHSS Link

Drone 4 Drone 3
FHSS Link

Figure 3. Diagram of the drone cloud with ring topology implementing FHSS with true random
frequency hopping.

6.1. Communication Infrastructure

In a ring topology, each drone has two communication links, one to the preceding
drone and another to the following one. This setup ensures redundancy and can facilitate
continuous data flow even if one link fails. The direct data links between drones make
efficient use of the available bandwidth and minimize latency.
Each data link operates independently with its own frequency hopping sequence.
This configuration helps in avoiding interference between links, and improves the overall
resilience of the network to jamming and interception attempts.

6.2. Frequency Hopping Implementation

Each drone is equipped with an RPU to generate true random frequency hopping
sequences. Given the ring topology, the sequence for each link can be independently
generated and is specific to the communicating pair of drones. The use of true random
numbers ensures that even if an adversary gains access to one link, they cannot predict the
hopping pattern of other links.
The true random numbers Ro generated by the RPUs are mapped to frequency values
f o within the communication band allocated for the drone cloud. The frequency for the o-th
interval is defined as: f o = f min + ( f max − f min ) RRmax
o
, where f min and f max are the minimum
and maximum frequencies available, and Rmax is the maximum output of the RPU.

6.3. Synchronization and Key Management

To establish communication, drones must initially synchronize their RPUs to ensure
that the frequency hopping sequences are correctly aligned. This synchronization can be
achieved using GPS time stamps or a dedicated synchronization signal.
Each drone pair must securely exchange an initial key or synchronization data to
coordinate the start of the frequency hopping sequence. Public-key cryptography can be
used for this purpose, providing a secure means of distributing the necessary information.
Future Internet 2024, 16, 412 18 of 33

Given the potential for clock drift or other synchronization issues, drones periodically
resynchronize their frequency hopping sequences. This can be achieved through a resynchro-
nization protocol, possibly using reserved channels or a periodic synchronization beacon.
While it is true that an initial secure and reliable channel is used for exchanging
keys and synchronization data between drones, the primary purpose of this channel is to
establish the parameters necessary for secure communication. However, relying solely on
this channel for all communication would expose the system to vulnerabilities associated
with fixed-frequency transmissions, such as jamming, interception, and eavesdropping.
The use of Frequency Hopping Spread Spectrum (FHSS) for the main communication
channel offers several critical advantages:
• Resistance to Jamming: FHSS systems are inherently resistant to jamming attacks
because the frequency changes rapidly in a pattern known only to the communicating
parties. An adversary attempting to jam the communication would need to jam all
possible frequencies simultaneously, which is often impractical.
• Enhanced Security: By hopping frequencies in a true random sequence, the com-
munication becomes extremely difficult to predict or intercept without knowledge
of the hopping pattern. This significantly reduces the risk of eavesdropping and
unauthorized access.
• Interference Mitigation: FHSS helps in mitigating interference from other devices
operating in the same frequency band. Since the communication frequency changes
rapidly, the impact of any interference on a particular frequency is minimized.
• Robustness in Adverse Environments: Drones often operate in environments where
the communication channel conditions can vary rapidly due to movement, obstacles,
or other factors. FHSS provides robustness against such variations, maintaining
reliable communication.
• Scalability: In a network of multiple drones, FHSS allows for better spectrum uti-
lization and reduces the likelihood of collisions, as different pairs can use different
hopping patterns.
The initial secure channel is typically a narrowband communication link used specif-
ically for key exchange and synchronization due to its simplicity and lower resource
requirements. It is not designed to handle the bandwidth and security requirements of
continuous drone communication, especially in hostile or dynamic environments. By
transitioning to an FHSS-based communication channel after the initial setup, the system
leverages the benefits of spread spectrum techniques to enhance overall communication
security and reliability.
Therefore, while the initial secure channel is essential for setting up communication
parameters, the use of FHSS for ongoing communication is critical for protecting against
threats and ensuring the robustness of the drone network in real-world operating conditions.

6.4. Complexity Analysis of Proposed Algorithm

The complexity of the proposed QRNG-based Frequency Hopping Spread Spectrum
(FHSS) algorithm is analyzed in terms of the random number generation, frequency hop-
ping, and synchronization.
• Random Number Generation: The QRNG generates a random number Ro per hop,
where each generation is effectively O(1) due to dedicated QRNG hardware.
• Frequency Selection: For each frequency hop, a lookup is performed to select f o from
Favail . Assuming the set Favail has size | F |, selecting a frequency f o involves a mapping
operation, making the complexity O(log | F |) for each hop due to binary search or
direct indexing in practice.
• Synchronization and Resynchronization: Each drone periodically synchronizes its
hopping sequence with the other drones in the network. Assuming N drones and
each drone communicates with two neighboring drones in a ring topology, the syn-
chronization process is O( N ).
Future Internet 2024, 16, 412 19 of 33

Given the above, the total complexity for frequency hopping across T time intervals in
a network of N drones is approximately:

O( T · (1 + log | F |) + N ) = O( T log | F | + N )
This represents a relatively low complexity, making the QRNG-based FHSS efficient
for real-time operation even in dense drone networks.

7. Evaluations
The primary objective of our experimentation was to evaluate the effectiveness and
security of using true random frequency hopping in a simulated drone network with a ring
topology. The setup included the following key components:
• QRNG Device or RPU: A Quside QRNG device [9,31] installed in the internal network
of the BARCELONA Supercomputing Centre Facility was utilized to generate true
random numbers, which were used to create the frequency hopping sequences. The
device is a QRNG Module ETH based on the FMC 400 Randomness Module by Quside.
• Drones: Each drone in the network was a simulated entity capable of implementing
frequency hopping based on the generated sequences.
• Secure Communication Protocol: A secure communication protocol was developed,
incorporating AES encryption, Hash-based message authentication code (HMAC) for
data integrity, and secure key exchange mechanisms.
• Synchronization Mechanism: Synchronization between drones was achieved using a
preset time, that in practice would correspond to GPS, ensuring a common start time
for frequency hopping across all nodes.
The QRNG Module ETH, see Figure 4, is based on the FMC 400 Randomness Module
by Quside. It includes an FPGA baseboard to provide high-quality and fast random bit
rates through a 1Gb-Ethernet connection.
The QRNG Module ETH employs a digital quantum entropy source based on the
accelerated phase diffusion process in semiconductor lasers. The system comprises:
• Laser: Produces a stream of phase-randomized optical pulses.
• Unbalanced MACH-ZEHNDER Interferometer: Converts random phases into ran-
dom amplitudes.
• Photodetector: Translates the optical signal into an electrical signal.
• Digitizer: Converts the analog signal into a digital bit stream.
• Control and Clocks: Manages the digital control, health monitoring, and random-
ness extraction.

Figure 4. Reference front panel for the QRNG Module ETH powered by the FMC 400 quantum
entropy source by Quside.

• RJ-45 Connection (Front Panel): Connects the QRNG Module ETH to the PC for
control, monitoring, and data transfer.
• ON/OFF Switch (Front Panel): Powers the device on or off. The system restarts with
default settings when switched on.
• Power Plug (Rear Panel): Connects to a 110/240 VAC power source.
• Status LED (Front Panel): Indicates the system status.
Future Internet 2024, 16, 412 20 of 33

This device leverages the inherent quantum phase noise of lasers to generate true
random numbers, providing a raw bit rate of up to 1 Gbit/s with strong entropy character-
istics. The phase diffusion technique involves modulating a semiconductor laser to produce
a stream of phase-randomized optical pulses, which are then converted into amplitude
fluctuations using an interferometer [32–34]. These fluctuations are detected and digitized
to produce high-quality random bit streams suitable for cryptographic applications and
secure communications.

7.1. Procedure
The experimentation procedure consisted of several stages:
The QRNG device generated true random numbers, which were converted into a
series of frequency values within a specified range. This sequence determined the order of
frequency changes for each communication link in the drone network.
To establish a secure communication channel, a shared key was exchanged between
the drones using a secure key exchange protocol. This key was used for both encrypting
the frequency sequence and generating HMACs for integrity verification.
All drones synchronized their internal clocks to a common preset time, which in
practice would correspond to a GPS-based time reference. An agreed-upon start time
was set, and all drones commenced frequency hopping simultaneously, adhering to the
pre-generated sequence.
During the simulation, each drone entity transmitted data packets over the network.
Here each drone is an instantation entity of its correspondent class in python. The data,
including the frequency sequence and synchronization information, was encrypted using
AES. HMACs were generated and attached to each packet to ensure data integrity.
The experiment’s performance was monitored in real-time, where the code in python
simulated a drone ring topology, and the QRNG was accessed through a lock mechanism
in the internal VPN of the BARCELONA Supercomputing Center. We focused on synchro-
nization accuracy, data integrity, and resistance to potential interception or jamming. The
frequency hopping sequence and synchronization data were logged for subsequent analysis.

7.2. Results and Observations

Figure 5 presents a comprehensive overview of our proposed methodology for en-
hancing the security and reliability of multi-drone communications using QRNG-based
FHSS systems. At the core of our approach is a QRNG device, which generates truly
random numbers to drive the frequency hopping sequences. These random numbers are
efficiently distributed to multiple drones through a shared memory system, ensuring that
each drone has access to high-quality random data for its FHSS operations. To manage
concurrent access to the QRNG device, we implement a specialized protocol that optimizes
resource utilization while maintaining the integrity of the random number generation
process. This architecture allows each drone to implement its own FHSS system using the
shared random numbers, resulting in a network of secure, jam-resistant communication
channels. The wireless medium, represented by the cloud in the diagram, emphasizes the
challenging environment in which these secure communications take place. By integrating
quantum randomness into a multi-drone FHSS framework, our methodology significantly
enhances the system’s resistance to jamming, interception, and other forms of interference,
thereby improving the overall security and reliability of drone communications in complex
operational scenarios.
Future Internet 2024, 16, 412 21 of 33

Drone 1
FHSS System

Distributes random numbers

Quantum Random
Shared Memory Drone 2 Wireless
Number Generator Secure communication
System FHSS System Channel
(QRNG)

Drone 3
FHSS System
Concurrent Access
Protocol
Manages access to QRNG

Figure 5. Diagram illustrating the methodology of using a QRNG for true random frequency hopping
in a multi-drone FHSS system. The QRNG generates truly random numbers, which are distributed to
multiple drones through a shared memory system. A concurrent access protocol manages access to
the QRNG, ensuring efficient use of the device. Each drone implements its own FHSS system using
the shared random numbers, enhancing anti-jamming capabilities and resistance to eavesdropping
across the entire network.

The experimental results, as illustrated in Figures 6–17, provide compelling evidence

of the effectiveness of our QRNG-based frequency hopping system across multiple drones.
The frequency distribution plots (Figures 7, 11 and 15) demonstrate a relatively uniform
spread of frequencies for all three drones, except for the presence of jamming at 5500 Hz,
indicating successful implementation of the quantum random number generation. The
frequency sequence plots (Figures 8, 12 and 16) show no discernible patterns, which is cru-
cial for maintaining unpredictability in the hopping sequence. Notably, the autocorrelation
plots (Figures 6, 10 and 14) exhibit sharp peaks at zero lag and linear decay to near-zero
values for all other lags, confirming a general absence of periodicity or predictability in the
sequences, albeit some short-term correlation; which indicates that the RPU technology has
still room for improvement in terms of quality and stability of the generated random num-
bers. The spectral analysis plots (Figures 9, 13 and 17) further corroborate these findings,
showing a relatively flat power spectral density across frequencies, which is characteristic
of truly random sequences. These results collectively demonstrate that our QRNG-based
system successfully generates unique, largely unpredictable, and uniformly distributed
frequency hopping sequences for each drone, thereby enhancing the security and reliability
of the communication system against potential jamming or interception attempts. That
being said, while QRNG technology is still at its early stage of development, it shows
promising performance in the future advancement of secure communication scenarios.
An important aspect of our experimental setup was the selection of parameters
to thoroughly evaluate the performance and resilience of our QRNG-based FHSS sys-
tem. Each drone generated a sequence of 10,000 frequency hops (num_hops = 10,000),
operating within a frequency range from 1000 Hz to 10,000 Hz (freq_min = 1000 Hz,
freq_max = 10,000 Hz). The duration of each frequency hop was fixed at 0.01 s
(duration = 0.01 s), resulting in a total communication time of 100 s per drone. To ensure
synchronized operations, all drones were programmed to commence their frequency hop-
ping sequences simultaneously, using a synchronization time set 10 s ahead of the current
system time (sync_time = current time + 10 s), as a proxy for GPS time.
The jamming simulation was implemented by introducing interference at a central
frequency of 5500 Hz. Frequencies within a 200 Hz range of this jamming frequency were
adjusted by adding a random offset between −500 Hz and 500 Hz, effectively simulating
an active jammer’s impact on the communication system. These parameters were chosen to
provide a test of the system’s anti-jamming capabilities, ensuring that the FHSS sequences
generated by the drones could adapt and maintain secure communications even in the
presence of targeted interference.
Future Internet 2024, 16, 412 22 of 33

Figure 6. Drone 0: Autocorrelation.

Figure 7. Drone 0: Frequency Distribution, where a noticeable dip around 5500 Hz indicates a
reduced frequency count in the jammed region.

A particular aspect of our implementation that warrants clarification is the method

by which random numbers from the QRNG were transformed into frequency hopping
sequences. We collected random bytes from the QRNG and assembled them into 32-bit
unsigned integers to maximize the range and uniformity of the random values. Each integer
was then normalized by dividing by 232 and scaled to fit within our desired frequency range.
This careful mapping was crucial to ensure a truly uniform distribution of frequencies,
eliminating any biases that could arise from improper normalization or scaling.
Future Internet 2024, 16, 412 23 of 33

Figure 8. Drone 0: Frequency Sequence over 100 hops of the total 100,000 hops.

Figure 9. Drone 0: Spectral Analysis.

The experimentation yielded several key insights:

• Data Security: The AES encryption and HMAC verification successfully protected the
frequency sequence and synchronization data from interception and tampering. No
unauthorized access or data modification was detected during the experiment.
• System Resilience: The implementation of true random frequency hopping demon-
strated a potential high degree of resilience to jamming, given by the spectral analysis
of the frequency hopping sequences. The unpredictability of the sequence made it
difficult for potential adversaries to disrupt communication.
Future Internet 2024, 16, 412 24 of 33

Figure 10. Drone 1: Autocorrelation.

Figure 11. Drone 1: Frequency Distribution, where a noticeable dip around 5500 Hz indicates a
reduced frequency count in the jammed region.

Although this is a simplified setup for illustration purposes, and given that QRNG
are still at their early stage of development and deployment, to address the practical
implementation of distributing true random numbers to drones that are airborne, we
could adapt the concept of a shared memory system to a distributed model suitable
for a flying multi-drone environment. Instead of relying on a physical shared memory,
which is not feasible for drones in flight, we would need to implement a virtual shared
memory system facilitated through secure wireless communication channels. The QRNG
device would be stationed at a central ground control unit or a leading drone acting as a
coordinator. This QRNG would generate true random numbers, which are then securely
transmitted to each drone in the network using encrypted communication links prior to
Future Internet 2024, 16, 412 25 of 33

or during the mission. Each drone maintains a local buffer to store the received random
numbers for its FHSS operations. To manage the distribution and synchronization of
the random numbers, we would need to implement a concurrent access protocol as a
network protocol that ensures all drones receive the necessary random data in a timely
and secure manner. This protocol handles requests for random numbers, manages the
transmission schedules to avoid collisions, and verifies the integrity of the received data
using cryptographic techniques such as digital signatures or message authentication codes
(MACs). By virtualizing the shared memory over the network, we enable each drone
to have access to high-quality random numbers without the need for a physical shared
memory. This approach makes our methodology practical and implementable for drones
that are flying, allowing each drone to independently implement its own FHSS system while
maintaining synchronization with the rest of the network. The use of secure communication
channels and robust synchronization protocols ensures that the system remains resistant
to jamming, interception, and other forms of interference, thereby enhancing the overall
security and reliability of drone communications in complex operational scenarios.

Figure 12. Drone 1: Frequency Sequence over 100 hops of the total 100,000 hops.

Additionally, we conducted Chi-squared and Kolmogorov-Smirnov (KS) tests on the

generated frequency sequences to assess the randomness and distribution uniformity across
the frequency spectrum. The test results for each drone are shown in Table 1.

Table 1. Qualitative Summary of Performance and Statistical Analysis for Each Drone.

Drone Statistical Analysis Performance Outcome Key Observations

Reliable communication Dynamic response to
Chi-squared: 263.59, p < 10−30
Drone 0 under interference frequency jamming; high
KS: 0.01704, p = 0.0053
at 5500 Hz resilience in message integrity
Stable communication Effective avoidance
Chi-squared: 240.11, p < 10−27
Drone 1 with reduced selection of jammed frequency
KS: 0.01572, p = 0.0107
of jammed frequencies maintaining data integrity
Consistent resilience to Demonstrates robustness
Chi-squared: 233.88, p < 10−25
Drone 2 interference, with effective and adaptive capabilities
KS: 0.01679, p = 0.0070
packet integrity verification in hostile environments
Future Internet 2024, 16, 412 26 of 33

Figure 13. Drone 1: Spectral Analysis.

Figure 14. Drone 2: Autocorrelation.

The results indicate a high level of randomness, with Chi-squared values suggesting
non-uniformity at a fine scale (in part due to active jamming interference introduced
at 5500 Hz). This impact aligns with the frequency distribution patterns observed in
each drone’s logs, where a slight dip occurs around the jammed frequency. The low
p-values for the KS tests indicate statistically significant deviations from perfect uniformity,
suggesting that QRNG technology, while highly effective in generating random sequences,
may still require further refinement to reduce short-term correlations in the frequency
hopping sequences.
Future Internet 2024, 16, 412 27 of 33

Figure 15. Drone 2: Frequency Distribution, where a noticeable dip around 5500 Hz indicates a
reduced frequency count in the jammed region.

Figure 16. Drone 2: Frequency Sequence over 100 hops of the total 100,000 hops.

The performance logs provide further insight into the stability and robustness of the
QRNG-based frequency hopping system. Drones 0, 1, and 2 each maintained reliable
communication despite jamming interference, with performance variations reflecting the
adaptive response capabilities of the frequency exclusion list management.
• Frequency Exclusion Management: Each drone demonstrated successful avoidance
of the jammed 5500 Hz frequency during the experiment, which is evidenced by
fewer frequency selections in the impacted range. This illustrates the system’s dy-
namic response to interference, essential for secure drone communication in hos-
tile environments.
• Transmission Integrity: Packet integrity checks indicated no unauthorized modifica-
tions or data corruption, verifying the encryption and message authentication proto-
cols in place. The HMAC verification success rate was consistent across all drones,
reinforcing the communication channel’s resilience under QRNG-driven FHSS.
Future Internet 2024, 16, 412 28 of 33

Figure 17. Drone 2: Spectral Analysis.

Overall, the combination of statistical tests and performance metrics provides a com-
prehensive assessment of the QRNG-based FHSS system’s reliability. These results confirm
its potential as a robust security solution for multi-drone networks, even in the presence
of environmental interference and adversarial actions. Further enhancements to QRNG
stability and random sequence generation are anticipated to improve performance in
future deployments.
To better position our work within the landscape of related studies, we provide a qual-
itative comparison table (Table 2). This table contrasts our proposed QRNG-based FHSS
method with other security approaches commonly employed in wireless communication
and drone networks. Key aspects compared include the type of security technique utilized,
each study’s adaptability to jamming attacks, the computational complexity, and the source
of randomness in frequency generation. This comparison underscores the advantages of us-
ing QRNG-based true random frequency hopping for enhanced anti-jamming adaptability
and security, particularly in applications where unpredictability is crucial.

Table 2. Qualitative Comparison with Related Studies.

Study Technique Anti-Jamming Adaptability Complexity Randomness Source

[23] Seamless UAV multicasting Medium (limited FHSS) O( N 2 ) for network transitions PRNG
[24] Jammer detection in IoT High (targeted detection) O( N log N ) PRNG
This Study QRNG-based FHSS for drones High (true random FHSS) O( N log | F |) for FHSS QRNG
[20] Secure routing for UAVs Medium (route-level resilience) O( N 2 ) for path finding PRNG
[21] Protocol optimization Low (limited hopping) O( N ) PRNG

7.3. Drone Experimental Configuration and Simulated Environment

The experimental setup for this study involved a simulated environment in Python to
test the FHSS with QRNG-generated sequences under controlled conditions. The simulation
environment involved the following configurations and conditions:
• Drone Configuration: Each drone in the network was represented by an instance of a
Python class, which implemented QRNG-based frequency hopping. Each simulated
drone had independent QRNG access through shared virtual memory, providing
real-time access to randomized frequency values. The drones communicated in a ring
topology, where each drone transmitted to two neighboring drones.
Future Internet 2024, 16, 412 29 of 33

• Communication Environment: The simulation modeled a wireless communication

channel subject to random interference (jamming) at specific frequency bands to
evaluate the system’s anti-jamming capabilities. Interference at a central frequency
(e.g., 5500 Hz) was introduced to observe the adaptability of the FHSS algorithm.
• Topology: A ring topology was chosen for its suitability for real-world drone networks,
where each drone maintains a secure line of communication with its immediate
neighbors. This configuration provides resilience and redundancy, as data can flow
continuously around the ring even if one link fails.
While tested in a simulated environment, this approach could be adapted to a real-
world drone network with ring topology. Each drone would require a lightweight QRNG
device or access to a ground-based QRNG for random number generation. The ring
configuration would enhance reliability and enable real-time data relay across the network,
even under jamming conditions.

8. Discussion
The use of true random frequency hopping is particularly valuable in military and
defense contexts, where secure and reliable communication is critical. In a drone cloud
configuration, such systems can be employed for reconnaissance, surveillance, and combat
support, ensuring that communications remain secure from interception and jamming. The
high security provided by true random sequences helps protect sensitive information and
command instructions.
Drones equipped with secure communication systems can play a vital role in disaster
response and search and rescue operations. In scenarios where the environment may be
hostile or communication infrastructure is compromised, a drone cloud using true random
frequency hopping can maintain secure and reliable lines of communication. This is crucial
for coordinating efforts, relaying information, and ensuring the safety of personnel involved
in the operations.
Industries such as agriculture, logistics, and infrastructure inspection increasingly
use drone swarms to collect data and perform tasks. True random frequency hopping can
secure the communication links between drones, preventing industrial espionage and data
theft. For instance, in precision agriculture, drones can securely communicate real-time
data on crop health and soil conditions, enhancing the decision-making process.
Drones are extensively used for environmental monitoring, including wildlife tracking,
pollution assessment, and natural resource management. Secure communication ensures
that sensitive environmental data is protected, and that the drones’ operations cannot
be easily disrupted. This is particularly important in protected or sensitive areas where
unauthorized access or interference could have significant ecological impacts.
In the context of smart cities, drones can be used for infrastructure monitoring, traffic
management, and public safety. True random frequency hopping ensures that these drones
can communicate securely, protecting data on critical infrastructure and avoiding potential
security breaches that could disrupt urban services.

Potential Future Developments

Future developments may involve integrating true random frequency hopping with AI
and machine learning algorithms to dynamically optimize communication strategies. For
example, AI could predict optimal frequency bands to use based on real-time environmental
data, enhancing the system’s efficiency and robustness.
The combination of true random frequency hopping with advanced encryption tech-
niques could further enhance communication security. This dual-layer security approach
could protect against both physical and digital threats, ensuring comprehensive protection
for communication systems.
The principles of true random generation could be extended into the realm of quan-
tum communication. Quantum key distribution (QKD), combined with true random
Future Internet 2024, 16, 412 30 of 33

frequency hopping, could provide unprecedented levels of security, leveraging the inherent
unpredictability of quantum states.
Future systems may focus on creating interoperable communication platforms that can
work seamlessly across different types of drones and other autonomous systems. This could
involve standardizing true random frequency hopping protocols to ensure compatibility
and secure communication between diverse systems.
Ongoing research will likely focus on miniaturizing RPUs and improving their power
efficiency. This is crucial for expanding the use of true random frequency hopping in
smaller drones and other devices with limited power resources. Advances in materials
science and semiconductor technologies could play a key role in these developments.
As drones become more integrated into critical infrastructure and services, they will
become targets for cyber-physical attacks. Future research will likely explore ways to make
true random frequency hopping systems resilient not only to traditional jamming and
interception but also to sophisticated cyber-attacks aimed at disrupting drone operations.

9. Conclusions
In this paper, we presented a novel approach to enhancing the security and reliability
of Frequency Hopping Spread Spectrum (FHSS) systems by integrating a Quantum Random
Number Generator (QRNG) to generate truly random frequency hopping sequences. Our
methodology was applied to a simulated multi-drone network with a ring topology, where
each drone utilized true random sequences for frequency hopping, significantly improving
resistance against jamming and interception attempts.
The experimental results provided strong evidence supporting the security and relia-
bility claims of our QRNG-based FHSS system. The statistical analyses, including spectral
density, frequency distribution, and autocorrelation studies of the generated frequency
sequences, demonstrated the high randomness and unpredictability achieved:
• Uniform Frequency Distribution: The frequency distribution analyses showed a rela-
tively uniform spread of frequencies across the available spectrum for all drones. De-
spite the presence of simulated jamming at 5500 Hz, the drones effectively minimized
frequency usage in the jammed region, indicating successful adaptation and resilience.
• Low Autocorrelation: The autocorrelation analyses revealed sharp peaks at zero lag
and near-zero values for other lags, confirming the absence of significant periodicity
or predictability in the sequences. This lack of correlation enhances resistance to
predictive attacks and jamming.
• Spectral Analysis: The spectral density plots exhibited a relatively flat power spectral
density across frequencies, characteristic of truly random sequences. This uniform
spectral distribution reduces vulnerabilities to spectral-based jamming and interception.
• Resilience to Jamming: The system demonstrated effective avoidance of the jammed
frequency region, maintaining high numbers of frequency hops in non-jammed re-
gions. This adaptability highlightsa the system’s robustness in maintaining communi-
cation integrity under interference.
• Statistical Randomness Tests: The Chi-squared and Kolmogorov-Smirnov tests per-
formed on the frequency sequences indicated high levels of randomness, with some de-
viations due to the simulated jamming. The test results confirm that the QRNG-generated
sequences exhibit the necessary randomness properties for secure FHSS operation.
The implementation of AES encryption and HMAC verification successfully pro-
tected the frequency sequences and synchronization data from interception and tampering.
Throughout the experiments, no unauthorized access or data modification was detected,
highlighting the effectiveness of the integrated security measures.
While the results are promising, some short-term correlations were observed in the
autocorrelation analyses, suggesting areas for improvement in QRNG technology. These
correlations, though minimal, indicate that the QRNG devices are still maturing and that
future enhancements could further improve the quality of the generated random numbers
and the stability of the device.
Future Internet 2024, 16, 412 31 of 33

The experimentation also highlighted practical considerations for deploying such

systems in real-world drone networks. The need for efficient distribution and synchro-
nization of true random numbers in airborne drones was addressed through the concept
of a virtual shared memory system facilitated by secure communication channels. This
approach enables each drone to access high-quality random numbers without the need for
physical shared memory, making the methodology practical for flying drones.
Building on these findings, future research will focus on:
• Optimizing Key Management and Synchronization: Enhancing the key management
processes and exploring advanced synchronization techniques to improve system
performance and security. This includes developing more efficient protocols for secure
random number distribution and synchronization among drones.
• Improving QRNG Technology: Testing upgraded versions of QRNG devices to address
the observed short-term correlations and improve the quality and stability of the
random number generation. Advancements in QRNG technology could lead to even
higher levels of randomness and system reliability.
• Real-World Implementation: Implementing the QRNG-based FHSS system in actual
drone networks to validate its performance in operational environments. This includes
assessing scalability, energy efficiency, and the system’s ability to handle real-world
interference and environmental factors.
• Integration with Advanced Security Measures: Combining the QRNG-based FHSS
system with other cryptographic techniques, such as quantum key distribution, to
further enhance communication security. This dual-layer security approach could
provide unprecedented levels of protection against both physical and digital threats.
• AI and Machine Learning Integration: Exploring the integration of AI and machine
learning algorithms to dynamically optimize communication strategies. AI could
be used to predict optimal frequency bands and adjust hopping patterns based on
real-time environmental data, enhancing efficiency and robustness.
In conclusion, the integration of QRNG technology into FHSS systems represents
a significant advancement in securing drone communications. The experimental results
substantiate the security and reliability claims of the proposed system, demonstrating
its effectiveness in mitigating potential threats such as jamming and interception. By
leveraging true randomness, the system achieves a higher level of unpredictability and
resilience, essential for secure communications in critical applications.
This work contributes to the growing field of quantum-enhanced wireless communica-
tions, offering substantial benefits for applications where enhanced security is paramount,
such as military operations, disaster response, and secure commercial drone activities. The
proposed methodology lays the groundwork for future developments in secure communi-
cation technologies, paving the way for more robust and reliable systems in increasingly
complex and adversarial environments.
Our findings confirm that QRNG-based FHSS systems have the potential to signifi-
cantly improve the security and reliability of drone communications. As QRNG technology
continues to mature, we anticipate further enhancements in system performance and
broader applicability across various domains requiring secure wireless communication.

Author Contributions: Conceptualization, J.d.C.; funding acquisition, C.T.C.; investigation, J.d.C.

and I.d.Z.; methodology, J.d.C. and I.d.Z.; software, J.d.C.; writing—original draft, J.d.C.; writing—
review and editing, J.d.C., I.d.Z., J.-C.C. and C.T.C. All authors have read and agreed to the published
version of the manuscript.
Funding: The research received the funding from the BARCELONA Supercomputing Center (TIFON).
Data Availability Statement: Data are contained within the article.
Acknowledgments: During the preparation of this work, the author(s) used ChatGPT/Copilot/Gemi-
ni/Claude for code completion and pseudocode generation, Grammarly for English language editing,
and Mendeley for bibliography management. The author(s) reviewed and edited the content as
Future Internet 2024, 16, 412 32 of 33

needed and take(s) full responsibility for the content of this publication. The authors would like to
thank the BARCELONA Supercomputing Center for providing access to MareNostrum 5 and tech-
nical support throughout this research. Thanks to the QUANTIC group, Artur García and Germán
Navarro, for discussions and instructional examples on the use of the QRNG device.
Conflicts of Interest: The authors declare that they have no conflict of interest. The funders had no
role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of
the manuscript, or in the decision to publish the results.

Abbreviations
The following abbreviations are used in this manuscript:

Quantum Random Number Generators QRNG

Randomness Processing Unit RPU
Pseudo-Random Number Generators PRNG
Frequency Hopping Spread Spectrum FHSS
Advanced Encryption Standard AES
Hash-based message authentication code HMAC
Elliptic Curve Cryptography ECC
Advanced Encryption Standard AES

References
1. de Zarzà, I.; de Curtò, J.; Roig, G.; Calafate, C.T. LLM Adaptive PID Control for B5G Truck Platooning Systems. Sensors 2023, 23,
5899. [CrossRef] [PubMed]
2. Dostert, K.M. Frequency-Hopping Spread-Spectrum Modulation for Digital Communications Over Electrical Power Lines. IEEE
J. Sel. Areas Commun. 1990, 8, 700–710. [CrossRef]
3. Liu, F.; Sun, G.; Zhang, S. Adaptive Measurement and Decoding of Frequency-Hopping Spread Spectrum Signals Based on
Knowledge Enhanced Compressed Sensing. IEEE Commun. Lett. 2021, 26, 1155–1159. [CrossRef]
4. Boutsioukis, N. Comparative Analysis of Pseudorandom Number Generators: Mersenne Twister, Middle Square Method, and
Linear Congruential Generator Through Dieharder Tests. 2023. Available online: https://ssrn.com/abstract=4354305 (accessed
on 1 September 2024).
5. Cannizzo, F. VMT19937: A SIMD-Friendly Pseudo Random Number Generator based on Mersenne Twister 19937. arXiv 2023,
arXiv:2309.16682. [CrossRef]
6. Liu, J.; Yao, X.; Wang, Y.; Wang, Z.; Xu, L.; Tong, H.; Liu, K. Novel Random SVPWM Technique Based on Tiny Mersenne Twister
PRNG to Reduce PWM Harmonic. In Proceedings of the Third International Conference on Mechanical Design and Simulation
(MDS 2023), Xi’an, China, 3–5 March 2023; SPIE Proceedings; Volume 12639, pp. 816–821. [CrossRef]
7. Soler, D.; Dafonte, C.; Fernández-Veiga, M.; Vilas, A.F.; Nóvoa, F.J. A privacy-preserving key transmission protocol to distribute
QRNG keys using zk-SNARKs. Comput. Netw. 2024, 242, 110259. [CrossRef]
8. Zheng, Z.; Guo, X.; Lin, F.; Wang, Y.; Wang, Y.; Guo, Y. Parallel CV-QRNG with Strict Entropy Evaluation. Photonics 2023, 10, 786.
[CrossRef]
9. Haider, Z.; Saeed, M.H.; Zaheer, M.E.; Alvi, Z.A.; Ilyas, M.; Nasreen, T.; Imran, M.; Islam, R.U.; Ikram, M. Quantum Random
Number Generator (QRNG): Theoretical and Experimental Investigations. Eur. Phys. J. Plus 2023, 138, 797. [CrossRef]
10. Pandey, S.K.; Jenef, R. A Comparative Study and Analysis of Quantum Random Number Generator with True Random Number
Generator. In Proceedings of the 2024 16th International Conference on COMmunication Systems & NETworkS (COMSNETS),
Bengaluru, India, 3–7 January 2024.
11. Lamarr, H.; Antheil, G. Secret Communication System. U.S. Patent No. 2,292,387, 11 August 1942.
12. Menezes, A.J.; van Oorschot, P.C.; Vanstone, S.A. Handbook of Applied Cryptography; CRC Press: Boca Raton, FL, USA, 1996.
13. Ma, X.; Yuan, X.; Cao, Z. Quantum random number generation. Npj Quantum Inf. 2016, 2, 16021. [CrossRef]
14. Jennewein, T.; Achleitner, U.; Weihs, G.; Weinfurter, H.; Zeilinger, A. A fast and compact quantum random number generator.
Rev. Sci. Instrum. 2000, 71, 1675–1680. [CrossRef]
15. Rivest, R.L.; Shamir, A.; Adleman, L. A method for obtaining digital signatures and public-key cryptosystems. Commun. ACM
1978, 21, 120–126. [CrossRef]
16. Miller, V.S. Use of elliptic curves in cryptography. In Advances in Cryptology—CRYPTO’85; Springer: Berlin/Heidelberg, Germany,
1985; pp. 417–426.
17. Koblitz, N. Elliptic curve cryptosystems. Math. Comput. 1987, 48, 203–209. [CrossRef]
18. National Institute of Standards and Technology (NIST). Announcing the Advanced Encryption Standard (AES); FIPS Publication 197;
NIST: Gaithersburg, MD, USA, 2001.
19. Abellán, C.; Amaya, W.; Jofre, M.; Curty, M.; Acín, A.; Capmany, J.; Pruneri, V.; Mitchell, M.W. Ultra-fast quantum randomness
generation by accelerated phase diffusion in a pulsed laser diode. Opt. Express 2014, 22, 1645–1654. [CrossRef] [PubMed]
Future Internet 2024, 16, 412 33 of 33

20. Gupta, L.; Jain, R.; Vaszkun, G. Survey of important issues in UAV communication networks. IEEE Commun. Surv. Tutor. 2016, 18,
1123–1152. [CrossRef]
21. Shin, D.; Sharma, V.; Kim, J.; Kwon, S.; You, I. Secure and Efficient Protocol for Route Optimization in PMIPv6-Based Smart Home
IoT Networks. IEEE Access 2017, 5, 11100–11117. [CrossRef]
22. Bekmezci, I.; Sahingoz, O.K.; Temel, Ş. Flying ad-hoc networks (FANETs): A survey. Ad Hoc Netw. 2013, 11, 1254–1270. [CrossRef]
23. Tu, W. Resource-Efficient Seamless Transitions for High-Performance Multi-Hop UAV Multicasting. Comput. Netw. 2022,
213, 109051. [CrossRef]
24. Abdollahi, M.; Malekinasab, K.; Tu, W.; Bag-Mohammadi, M. Physical-Layer Jammer Detection in Multihop IoT Networks. IEEE
Internet Things J. 2023, 10, 20574–20585. [CrossRef]
25. Zhu, J.; Wang, A.; Wu, W.; Zhao, Z.; Xu, Y.; Lei, R.; Yue, K. Deep-Learning-Based Recovery of Frequency-Hopping Sequences for
Anti-Jamming Applications. Electronics 2023, 12, 496. [CrossRef]
26. Eltholth, A.A. Improved Spectrum Coexistence in 2.4 GHz ISM Band Using Optimized Chaotic Frequency Hopping for Wi-Fi and
Bluetooth Signals. Sensors 2023, 23, 5183. [CrossRef]
27. Sokolov, V.; Skladannyi, P.; Platonenko, A. Jump-Stay Jamming Attack on Wi-Fi Systems. In Proceedings of the 2023 IEEE
18th International Conference on Computer Science and Information Technologies (CSIT), Lviv, Ukraine, 19–21 October 2023;
pp. 192–195.. [CrossRef]
28. Pärlin, K.; Riihonen, T.; Le Nir, V.; Adrat, M. Physical-layer reliability of drones and their counter-measures: Full vs. half duplex.
IEEE Trans. Wirel. Commun. 2023, 22, 5566–5581. [CrossRef]
29. Ganesan, T.; Jayarajan, N.; Shri Varun, B.G. Dynamic Control, Architecture, and Communication Protocol for Swarm Unmanned
Aerial Vehicles. In Computing in Intelligent Transportation Systems; Springer International Publishing: Cham, Switzerland, 2023;
pp. 31–49.
30. Herbst, J.; Müller, R.; Lipps, C.; Schotten, H.D. A Ring Topology Approach: Efficient Communication in Wireless Body Area
Networks (WBANs). In Proceedings of the 2024 Joint European Conference on Networks and Communications & 6G Summit
(EuCNC/6G Summit), Antwerp, Belgium, 3–6 June 2024; pp. 1084–1089.
31. Cirauqui, D.; García-March, M.Á.; Guigó Corominas, G.; Graß, T.; Grzybowski, P.R.; Muñoz-Gil, G.; Lewenstein, M. Comparing
pseudo- and quantum-random number generators with Monte Carlo simulations. APL Quantum 2024, 1, 036125. [CrossRef]
32. Mitchell, M.W.; Abellán, C.; Amaya, W. Strong Experimental Guarantees in Ultrafast Quantum Random Number Generation.
Phys. Rev. A 2015, 91, 012314. [CrossRef]
33. Álvarez, J.; Sarmiento, S.; Lázaro, J.; Gené, J.; Torres, J. Random Number Generation by Coherent Detection of Quantum Phase
Noise. Opt. Express 2020, 28, 5538–5547. [CrossRef] [PubMed]
34. Imran, M.; Sorianello, V.; Fresi, F.; Jalil, B.; Romagnoli, M.; Potì, L. On-Chip Tunable SOI Interferometer for Quantum Random
Number Generation Based on Phase Diffusion in Lasers. Opt. Commun. 2021, 485, 126736. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.