Project Proposal

MSc in Artificial Intelligence

Software Defined Radio Based Intelligent System for Modulation

G.A.S.Y Nanayakkara
208538X

Faculty of Information Technology

2022

0
 Project Proposal

MSc in Artificial Intelligence

Software Defined Radio Based Intelligent System for Modulation

G.A.S.Y Nanayakkara
208538X

Faculty of Information Technology

2022

1
x
 Contents

Page
1. Introduction 1
2. Aim & Objectives 2
3. Literature Review 2
4. Problem in Brief 3
5. Proposed Solution 3
6. Resource Requirements 4
7. Summary 4

References

Appendix - Plan of Action 6

i
0
1. Introduction
Wireless services and apps have grown pervasive in many aspects of people's life. Wireless
technologies are used not just for smart phones, but also for telemetry, surveillance, emitter
location, radio navigation, jamming, anti-jamming, radar detection, Unmanned Aerial
Vehicle (UAV) surveillance, navigation, and location tracking. With such widespread
reliance on the radio frequency (RF) spectrum, it is critical that people manage and use the
limited available spectrum as efficiently as possible. At the same time, electromagnetic
spectrum has become very congested because to the rising domain and widespread usage of
RF transmitters.
Electronic warfare is a military activity that uses the electromagnetic spectrum to attack an
enemy or impede enemy assaults, and jamming is a form of electronic warfare where
jammers radiate interference signals toward an enemy RF signals such as communication,
self-attacking drones, surveillance drones and etc. in order to disrupt the receiver. Airborne
drone detection is critical for interdiction. This is divided into four sections: Radio Detection
and Ranging (RADAR), optical sensors such as day and night vision cameras, acoustic
sensors and RF analysis [1]. RF analysis is not yet being employed by Sri Lankan armed
personnel; instead, only optical sensors and RADAR are being used to detect drones.
However, in a situation where an aerial assault is possible, analytical decision-making is
crucial. Hence, RF analysis are very much important because it keeps an eye on the RF
spectrum and finds the signals used to operate drones.
To combat security vulnerabilities posed by rogue or unknown transmitters, RF transmitters
should be detected not only by the data content of broadcasts but also by the basic physical
properties of the transmitters. Due to the extraordinarily high data rates involved and the
potential for several transmitters to share a channel in one area, RF wave forms pose a unique
difficulty [2]. These characteristics highlight the necessity for faster fingerprinting and
identifying procedures that go beyond traditional hand-engineered methods [3]. In this work,
RF data from the drone remote is identified and collected using Software Defined Radio
(SDR), a radio that employs software to do signal-processing tasks that were previously
accomplished by hardware. Then, as a counter-attack, a machine learning model will be
provided to train and classify modulation strategies utilized in drone communication, as well
as a suitable jamming strategy.
Presently I’m serving as a Research & Development Engineer at Sri Lanka Naval Dockyard
and possess the resources of obtaining several drone transmitter modulated signals using
SDR as data to design the system.
Therefore, in this research, an intelligent system is proposed for modulation classification
and jamming of drones based on SDR. The rest of the proposal has been structured with
background and motivation, research problem, proposed solution, resource requirement and
the summary.

1
2. Aim & Objectives
Following objectives have been identified to develop an intelligent system for modulation
classification and jamming of drones with the use of SDR.
Objective 1 - Critical review of literature in different drone transmitters and specific
modulation techniques.
Objective 2 - In depth study of the SDR and related Machine Learning / Deep Learning
approaches.
Objective 3 - Design and develop prototype of the software and hardware based jammer.
Objective 4 - Evaluate the prototype.
Objective 5 - Preparation of thesis and paper for conference.

3. Literature Review
UAVs have recently been popular in a variety of military, civil, agriculture and scientific
applications such as climate monitoring, crisis management, item/food delivery, search and
rescue operations, space exploration, and animal tracking [4]–[7]. According to [8], the UAV
industry would rise from USD 27.4 billion in 2021 to USD 58.4 billion by 2026. This
anticipated rise is mostly due to rising demand for automation and quick developments in
supporting technology. However, such technology is also employed in acts of damage like
hostile attacks in addition to being used for human advancement.
Several case studies have been identified as per [1], regarding drone attacks around the world
and consequences are discussed as follows. The sarin gas, one of the most lethal chemical
warfare weapons known, was intended to be sprayed using a remote-controlled helicopter by
Aum Shinrikyo in 1994, but testing failed when the chopper crashed. A planned Al-Qaeda
assault in Pakistan employing many drones was foiled by local law enforcement in 2013. At
the same year in 2013, a targeted attack on a power distribution plant in California nearly
knocked out electricity to a large area of the state. The damage suffered by the attack on this
unmanned plant was estimated to be $15 million USD. By 2014, the Islamic State has begun
to use commercial and homemade aerial drones against the troops in Iraq and Syria. A swarm
of 13 homemade aerial drones assaulted two Russian military outposts in Syria in January
2018. Furthermore, Russia claims that its Black Sea navy headquarters in Sevastopol was
targeted by a Ukrainian drone strike, injuring five people and forcing authorities to cancel
Navy Day celebrations [9]. Terrorist organizations have employed airborne drones for a
variety of operations, including information gathering, the delivery of explosives (either by
dropping explosives or using a drone outfitted with a rocket launcher), and the delivery of
chemical weapons. The deployment of chemical and biological weapons is a particularly
ominous use for drones. Hence, the awareness by detection as well as analysis will be much
more important when it comes to the decision making during such scenarios.
Optical detection refers to the employment of video cameras and computer algorithms in
drone detecting systems. These systems have limits due to low light conditions or weather
interference, as well as a high rate of false alarms. Low-flying and tiny drones can be found
using radar, although high levels of clutter can make this challenging. The majority of radar
systems are unable to distinguish between tiny drones and birds. Built-in software

2
restrictions, such as geofencing for drones with GPS capabilities, are the primary point of
failure for the majority of commercially available systems [9]. But still it is possible to
bypass the geofencing with some skillful personal. Therefore, in these circumstances, RF
analysis and the introduction of defense mechanisms such supporting effective RF jamming
techniques would be crucial.
The communication between the drone and the pilot or GPS is hindered by RF jamming. The
jamming techniques are basically classified as noise jamming, range deception jamming,
velocity deception jamming, and angle deception jamming [10]. Because the effect of
jamming varies based on the parameters of the received drone transmitter signal, an effective
jamming approach for the threat signal must be used. When a threat signal is received, the
traditional approach for determining a jamming strategy is based on information stored in a
library that maintains the suitable jamming method for signal kinds. There is a restriction to
the usage of a library when a threat signal of a new type or one that has been changed
differently from existing kinds is received. While a jamming technique can be simply chosen
and implemented when the type and value of a received threat signal exist in the library.
Therefore, it is required to use a machine learning model that, by learning the characteristics
of the received drone transmitter signal, can forecast the suitable jamming approach [11] .
Furthermore, jamming techniques such as barrage jamming may be utilized by a single
jammer to jam numerous frequencies at the same time [12]. The primary disadvantage of this
strategy is that the jammer spreads its strength across numerous frequencies, making it less
strong at a single frequency and potentially disrupting other military communications [13].
4. Problem in Brief
The problem statement is the unavailability of an intelligent system that could automatically
detecting and classifying modulation strategies used in drone communication and associated
transmitter types, as well as relevant counter-attack mechanisms.
5. Proposed Solution
An intelligent system is proposed to detect and classify RF modulation techniques of drones,
identification of related transmitter and counter attack based on appropriate jamming
technique. In order to generate the RF data, commercially available drone transmitters are to
be used. With the help of SDR, generated data will be recorded through receive signal path as
proposed system block diagram in Figure 1, and labeled according to the transmitter type and
modulation technique. Then the data will be preprocessed and several types of ML models
are to be created to train the data set. Further most accurate model will be selected and testing
is to be done by creating the real-time scenarios and automatic repeat attack will be created
accordingly through transmit signal path.

3
 Figure 1: Functional Block Diagram of the proposed system

The hypothesis of this project is that the intelligent decision-making while automatic RF
spectrum analysis, speed modulation identification and classification without any human
involvement, during a drone attack based on AI solution with Machine Learning. The
inspiration behind this hypothesis comes from evidence of previous incidence and power of
SDR. The proposed solution produces the output as an intelligent solution for drone
jamming. This decision is determined by several parameters including transmitting power,
Signal to Noise Ratio (SNR), IQ sampling, modulation technique and features of transmitter
and receiver.

The proposed solution under goes through many steps to generate the output from input,
which contains many parameters including automatic RF spectrum analysis and modulation
classification. This solution comprises two major roles as building the ML solution, as well
as design of software and hardware based jammer.

6. Resource Requirement
Following equipment, resources and software will be utilized in the project;

• HackRF 01/HDSDR module as SDR for signal processing

7. Summary
This proposal presented a novel system to identify and classify possible drone attacks based
on their controlled transmitter, RF modulation technique with the help of SDR based
intelligent system and further counter attack automatically by using appropriate jamming
technique. The proposed method will be modeled and tested by using commercially available
drone transmitters for evaluation. This solution will overcome the drawbacks of the
conventional drone identification systems such as optical sensing and RADAR.

4
References
[1] B. D. Joseph, “Weaponised drones: an airborne threat,” Daily News.
https://www.dailynews.lk/2022/01/11/features/269849/weaponised-drones-airborne-
threat (accessed Aug. 07, 2022).
[2] K. Youssef, L.-S. Bouchard, K. Z. Haigh, H. Krovi, J. Silovsky, and C. P. V. Valk,
“Machine Learning Approach to RF Transmitter Identification.” arXiv, Nov. 07, 2017.
doi: 10.48550/arXiv.1711.01559.
[3] S. Ponnaluru and S. Penke, “A software-defined radio testbed for deep learning-based
automatic modulation classification,” Int. J. Commun. Syst., vol. 33, no. 15, p. e4556,
2020, doi: 10.1002/dac.4556.
[4] A. Bhardwaj, L. Sam, Akanksha, F. J. Martín-Torres, and R. Kumar, “UAVs as remote
sensing platform in glaciology: Present applications and future prospects,” Remote Sens.
Environ., vol. 175, pp. 196–204, Mar. 2016, doi: 10.1016/j.rse.2015.12.029.
[5] J. Qi et al., “Search and Rescue Rotary-Wing UAV and Its Application to the Lushan Ms
7.0 Earthquake,” J. Field Robot., vol. 33, no. 3, pp. 290–321, 2016, doi:
10.1002/rob.21615.
[6] M. Messinger and M. Silman, “Unmanned aerial vehicles for the assessment and
monitoring of environmental contamination: An example from coal ash spills,” Environ.
Pollut., vol. 218, pp. 889–894, Nov. 2016, doi: 10.1016/j.envpol.2016.08.019.
[7] R. S. Allison, J. M. Johnston, G. Craig, and S. Jennings, “Airborne Optical and Thermal
Remote Sensing for Wildfire Detection and Monitoring,” Sensors, vol. 16, no. 8, Art. no.
8, Aug. 2016, doi: 10.3390/s16081310.
[8] “Unmanned Aerial Vehicle (UAV) Market Size, share | 2021-2026.”
https://www.marketsandmarkets.com/Market-Reports/unmanned-aerial-vehicles-uav-
market-662.html (accessed Aug. 07, 2022).
[9] A. Roth, I. Koshiw, and P. Sauer, “Russia claims five injured in Ukraine drone attack on
Black Sea fleet HQ,” The Guardian, Jul. 31, 2022. Accessed: Aug. 07, 2022. [Online].
Available: https://www.theguardian.com/world/2022/jul/31/russia-claims-ukraine-drone-
attack-black-sea-fleet-headquarters
[10] “Radar Vulnerability to Jamming. R. L. Lothes, M. B. Szymanski and R. G. Wiley. 247
pages, 23·5 × 15·5 cm, Artech House, Boston, 1990. £49. | The Journal of Navigation |
Cambridge Core.” https://www.cambridge.org/core/journals/journal-of-
navigation/article/abs/radar-vulnerability-to-jamming-r-l-lothes-m-b-szymanski-and-r-g-
wiley-247-pages-235-155-cm-artech-house-boston-1990-
49/B296572D048533CE35B37D9905EF6315 (accessed Aug. 10, 2022).
[11] G.-H. Lee, J. Jo, and C. H. Park, “Jamming Prediction for Radar Signals Using Machine
Learning Methods,” Secur. Commun. Netw., vol. 2020, p. e2151570, Jan. 2020, doi:
10.1155/2020/2151570.
[12] “Jamming Detection and Classification in OFDM-Based UAVs via Feature- and
Spectrogram-Tailored Machine Learning | IEEE Journals & Magazine | IEEE Xplore.”
https://ieeexplore.ieee.org/document/9707819 (accessed Aug. 07, 2022).
[13] “An Introduction to Jammers and Jamming Techniques - JEM Engineering.”
https://jemengineering.com/blog-an-introduction-to-jammers/ (accessed Aug. 07, 2022).

5
 Appendix

Plan of Action

Action Time in Months

Critical review of literature in different drone 01/08/2022 - 01/04/2023

Writing the thesis and paper for conference. 10/08/2022 - 22/04/2023