Detecting Jamming Attacks in Wireless Communication Using

Machine Learning Models

Paradesi Subba Rao, Nooka Varsha Reddy, Shaik Suhani, Kasetty Susmitha,

Lalam Jhansi and Pinjari Aneefa

Department of Computer Science and Engineering, Santhiram Engineering College, Nandyal‑518501, Andhra Pradesh,

India

Keywords: Jamming Detection, Machine Learning, RSSI, SNR, BER, Packet Loss Rate, Conventional, Software‑Defined

Radio, Threshold.

Abstract: Attacks from indiscriminate jammers could prevent communications by disrupting wireless networks. In

conventional jamming detection systems, software-defined radios or fixed-threshold signal evaluation

algorithms are incorporated in an attempt to solve the problem. These methods do not cope well with

sophisticated and adaptive jamming techniques due to inflexibility, excessive resource expenditure, and high

rates of false alarms. Fixed-threshold techniques cannot be adjusted adaptively, while methods based on SDR

necessitate high volumes of both processing power and costly radio frequency hardware. Machine learning

algorithms based jamming detection systems take features such as RSSI, SNR, BER, and packet loss rate as

detection model metrics. The proposed solution enhances the robustness and security of wireless

communication systems by providing fast, adaptable and hardware-independent response to jamming

inquiries.

1 INTRODUCTION

First of all, 5G, the Internet of Things, self-operation

systems and defense operations depend on this

technology of wireless communication. However,

such networks are vulnerable to interference attacks,

where a malicious person disrupts communication

to block transmission. These attacks can cause

serious security issues, loss of data and disruption of

networks, thus requiring effective detection

techniques.

More traditional jamming detection techniques

use Software-Defined Radios (SDRs), or fixed-

thresholds. Even with existing threshold-based

approaches monitoring network metrics (e.g., RSSI,

SNR, BER, packet loss rates), there are many false

positive events due to these disturbances.

Nonetheless, while the SDR-based method yields

accurate and precise results, the practical

implementation is limited by high computational

requirements and expensive hardware. In order to

cover these limitations, in this paper we propose a

machine learning-based jamming detection system.

The approach increases detection accuracy and

eliminates the need for additional devices.

Using Random Forest, Support Vector Machines

(SVM), Neural Networks, the proposed system

characterizes network states as normal or jammed in

real-time. Because the model continuously monitors

the actions in the network, it is highly scalable,

effective, and affordable, helping to discover new

kinds of jamming techniques and ensuring that

wireless networks work in a secure manner. The

optimal solution is applicable to various wireless

communication situations, ranging from 5G

networks, IoT ecosystems, and military defence

applications to critical infrastructure for its

convenience, effectiveness, and cost implications.

2 LITERATURE REVIEW

2.1 Machine Learning

Machine learning considers learning as a computer

science activity or task. In contrast to traditional

Rao, P. S., Reddy, N. V., Suhani, S., Susmitha, K., Jhansi, L. and Aneefa, P.

Detecting Jamming Attacks in Wireless Communication Using Machine Learning Models.

DOI: 10.5220/0013891700004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 3, pages

63-66

ISBN: 978-989-758-777-1

programming, ML does not restrict itself to the

program specifying a sequence of steps. In this way,

"gathers" means that it learns to predict by analyzing

data and identifying patterns. In supervised (the

models are trained on labeled data), unsupervised

(the data themselves are exploited to learn patterns,

etc. without a priori assigned labels), and

reinforcement (learning the decision-making process

from trial-by-trial experience) ML it is also feasible

to partition the ML in various sub-classes. Speech

recognition, fraud detection, recommender systems,

and predictive analytics are among these. Computing

lies in the large, databases being updated and scored

continuously as a stream of new data comes in, an

important feature of businesses in the fields of

cybersecurity, health, and automation.

2.2 How Machine Learning Is Used in

Detecting in Jamming Attacks

Jamming attack detection over wireless

communications is claimed to be one of the most

difficult problems, as the attacker uses an adaptive

interference recovery that is more complex than the

standard assumptions to detect ordinary jammers.

These types of attacks introduce radio frequency

interference into the input channel to compromise or

stop communication. This results in a loss or

stopping of communication services, which in turn

causes network and service disruptions.

Machine learning (ML) offers a highly effective

method, as it can track network activities, learn from

the past, and identify malicious activities that

indicate jamming. The algorithms, based on machine

learning, can be trained to identify the attack

signatures in the network parameters, e.g., Received

Signal Strength Indicator (RSSI), Signal-to-Noise

Ratio (SNR), Bit Error Rate (BER), and Packet Loss

Rate. Jamming-type attacks usually cause substantial

packet loss, an increase in BER, and a sudden

decrease in RSSI, all of which the ML models are

trained to associate with jamming.

3 METHODOLOGY

This section describes the methodology used to

identify jamming attacks; in particular, section 3.1

describes the project architecture (figure 1), section

3.2 describes the dataset information, section 3.3

describes the feature extraction approach.

3.1 Architecture

Figure 1: System Architecture.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

3.1 Dataset Information

RSSI-Measures the strength of incoming signal in

dBm (decibel-milliwatts), lower value represents

weaker signals.

SNR - The measurement of the ratio of signal power

to noise power in decibels. Higher values indicate

better signal quality.

BER - Describes the percent of bits that were routed

incorrectly due to some form of destruction. Greater

number of errors means higher figure of BER.

Packet Loss Rate (%) - Percentage of packets that did

not reach their destination, indicative of disturb at the

network.

Label - Defines the state of network set to be

Jamming (1) or Normal (0).

3.2 Feature Extraction

F1: RSSI analyses normal operations that are

characterized by higher RSSI values, whereas lower

values signal jamming activity.

F2: SNR analyses Elevated SNR signifies positive

network conditions, while lower SNR denotes

potential jamming.

F3: BER Analysis the interference is likely with high

BER values, while low BER values indicate stable

communications.

F4: Packet loss rate Normal operations are

characterized by lower packet loss, whereas higher

packet loss indicates jamming activity.

F5: Jamming attack classification during training the

model, classify network states into 'normal' (0) or

'jamming' (1) for use in the model.

4 IMPLEMENTATION AND

RESULTS

In this section the implementation details are

mentioned to detect the jamming attacks. 4.1 Section

contains the model selection and 4.2 section contains

the results of the implements.

4.1 Model Selection

Model 1: Random Forest

Random Forest is a strong machine learning model

which predicts results by averaging a few decision

trees. Since every individual tree inspects other

segments of the data, the final model improves on

their accuracy and strength. The combined power

helps it to address various parameters like RSSI,

SNR, BER, and Packet Loss Rate making it more

useful for jamming threat identification.

Model-2: Support Vector Machine

In SVM we train the data of this type of model

which differentiate between the data which

experiences interference in the network and that

which works normal. Under the condition that there

are clean features that can significantly separate

normal signals from jamming attacks, performance

of the technique is at its best. While SVM is

relatively slow on large datasets, it is an invaluable

approach when working with high-dimensional data.

Model-3: K-Nearest-Neighbour

KNN estimates network states based on

neighbouring data points. If most of the closest points

have been marked as normal based on network

conditions, the new point is assigned normal; if the

area is congested, the new point is considered

jammed.

Model-4: Decision Tree Classifier

A Decision Tree Classifier is a simple and

interpretable machine learning method for

classification problems. Its decision rules create a

tree-like structure, where the data is partitioned into

branches based on certain features.

Model-5: Logistic Regression

The Logistic Regression model was a basic model

which looked at what the probability was that the

network was working correctly versus congested.

Simple to use, efficient in operation, but less

effective than other models like Random Forest or

Neural Networks in terms of complex jamming

attack pattern handling.

Model-6: Multi-Layer Perceptron

The MLP is just yet another neural network that is

aimed at classifying and identifying through

patterns. The architecture includes one input layer,

one or more hidden layers, and one final output layer.

As the computational weight of each neuron, it

carries its calculations through a function that takes

the incoming messages and sorts through them to

filter and enhance the respective elements, such as

through use of Rectified Linear Units (ReLU) or a

Sigmoid Functions. Therefore, since MLP is

adaptive to learn from RSSI, SNR, BER, and Packet

Loss Rate, one of the useful tools for learning

adaptively from the data of the inherent complex

patterns of these metrics is the detection of jamming

attacks.

Detecting Jamming Attacks in Wireless Communication Using Machine Learning Models

4.2 Results

ROC-AUC curve as shown in figure 2 is a pictorial

representation of the true positive rate versus the

false positive rate.

Figure 2: ROC-AUC Scores of Test Data.

5 CONCLUSIONS

In our paper we provide a solution for jamming

attacks using machine learning models. Out of the

models tested, the Random Forest Classifier proved

to be the best performer with a test accuracy of 97%

and a ROC-AUC score of 99.01%. The performance

metrics demonstrate how well our method works to

differentiate between normal network traffic and

jamming attacks, which would significantly increase

the security of wireless networks.

REFERENCES

Chaitanya, V. Lakshmi, and G. Vijaya Bhaskar. "Apriori vs

Genetic algorithms for Identifying Frequent Item Sets."

International journal of Innovative Research

&Development 3.6 (2014): 249-254.

Chaitanya, V. Lakshmi. "Machine Learning Based

Predictive Model for Data Fusion Based Intruder Alert

System." journal of algebraic statistics 13.2 (2022):

2477-2483

Chaitanya, V. Lakshmi, et al. "Identification of traffic sign

boards and voice assistance system for driving." AIP

Conference Proceedings. Vol. 3028. No. 1. AIP

Publishing, 2024

Devi, M. Sharmila, et al. "Machine Learning Based

Classification and Clustering Analysis of Efficiency of

Exercise against Covid-19 Infection." Journal of

Algebraic Statistics 13.3 (2022): 112-117.

Devi, M. Sharmila, et al. "Extracting and Analyzing

Features in Natural Language Processing for Deep

Learning with English Language." Journal of Research

Publication and Reviews 4.4 (2023): 497-502.

Mahammad, Farooq Sunar, Karthik Balasubramanian, and

T. Sudhakar Babu. "A comprehensive research on

video imaging techniques." All Open Access, Bronze

(2019).

Mahammad, Farooq Sunar, and V. Madhu Viswanatham.

"Performance analysis of data compression algorithms

for heterogeneous architecture through parallel approa

ch." The Journal of Supercomputing 76.4 (2020): 2275-

2288.

Mahammad, Farooq Sunar, et al. "Key distribution scheme

for preventing key reinstallation attack in wireless

networks." AIP Conference Proceedings. Vol. 3028.

No. 1. AIP Publishing, 2024.

Mandalapu, Sharmila Devi, et al. "Rainfall prediction

using machine learning." AIP Conference Proceedings

. Vol. 3028. No. 1. AIP Publishing, 2024.

Mr.M.Amareswara Kumar, “Baby care warning system

based on IoT and GSM to prevent leaving a child in a

parked car” in International Conference on Emerging

Trends in Electronics and Communication Engineerin

g - 2023, API Proceedings July-2024

Mr.M.Amareswara Kumar, “Effective Feature Engineerin

g Technique for Heart Disease Prediction with Machine

Learning” in International Journal of Engineering &

Science Research, Volume 14, Issue 2, April-2024 with

ISSN 2277-2685.

Paradesi SubbaRao,”Morphed Image Detection using

Structural Similarity Index Measure”M6 Volume 48

Issue 4 (December 2024),https://powertechjournal.co

Paradesi Subba Rao, ”Detecting malicious Twitter bots

using machine learning” AIP Conf. Proc. 3028, 020073

(2024),https://doi.org/10.1063/5.0212693

Parumanchala Bhaskar, et al. "Machine Learning Based

Predictive Model for Closed Loop Air Filtering

System." Journal of Algebraic Statistics 13.3 (2022):

416-423.

Parumanchala Bhaskar, et al. "Incorporating Deep Learning

Techniques to Estimate the Damage of Cars During the

Accidents" AIP Conference Proceedings. Vol. 3028.

No. 1. AIP Publishing, 2024.

Parumanchala Bhaskar, et al “Cloud Computing Network

in Remote Sensing-Based Climate Detection Using

Machine Learning Algorithms” remote sensing in earth

systems sciences (springer).

Suman, Jami Venkata, et al. "Leveraging natural language

processing in conversational AI agents to improve

healthcare security." Conversational Artificial

Intelligence (2024): 699-711.

Sunar, Mahammad Farooq, and V. Madhu Viswanatham.

"A fast approach to encrypt and decrypt video streams

for secure channel transmission." World Review of

Science, Technology and Sustainable Development

14.1 (2018): 11-28.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES