Scribd
Upload
5 views

Presentation 11

Uploaded by

Sana Khalid
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
5 views

Presentation 11

Uploaded by

Sana Khalid
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

Topic: DDOS Attack

Submitted by: Group_12


Submitted to: Ma'am Ayesha
Deep learning
in distributed
denial-of • Authors: Firas Mohammed Aswad, Ali
service Mohammed Saleh Ahmed, Nafea Ali
attacks Majeed Alhamdi, Bashar Ahmad Khalaf,
and Salama A. Mostafa.
detection
method for • Published : 04/JULY/2022
Internet of
Things
networks
DDOS ATTACK
Major DDOS Attacks in History
IOT (Internet of Things)
Main Purpose of This Research
Deep Learning Concept
Outline: Previous Research Work for DDOS attacks detection
Research Gap in Previous Work
Work Done in this Research
Advantages
Disadvantages(Research Gap)
DDOS ATTACKs

DDoS Attack means "Distributed Denial-of-Service (DDoS) Attack " and it is a cybercrime in which the
attacker floods a server with internet traffic to prevent users from accessing connected online services and
sites.

Initially DDOS attack is launched by using only one system.

First major DDOS attack was launched by a high school student Michael Calce.". He targeted prominent
websites like amazon, yahoo and Dell etc.

The Dyn DDOS attack :Mirai Botnet attack which occurred in 2016 .
The DYN
ATTACK
IOT(Internet of Things)

• IOT means physical objects connected with each other and with the
Internet and exchanging data without human intervention.
• According to Cisco estimate,50 billion devices are now connected to
internet. But devices are fundamentally resource- restricted.
• Smart homes, Mobiles ,Watches and Refrigerators.
Major Purpose Of This Research

• Main purpose of this research was to protect the IOT devices and
controlling the DDOS attack using deep learning algorithms. Its main
purpose was to detect DDOS attacks without affecting legitimate traffic
over a network. This research tried to propose a model which can be used
as a standard to detect DDOS attacks.
DEEP Learning
• Deep learning is a type of machine learning that uses neural networks to analyze data and
draw conclusions like humans.
• Deep learning is a method that trains computers to process information in a way that
mimics human neural processes.
• Deep learning is a branch of machine learning that is made up of a neural network with three
or more layers:
• Input layer: Data enters through the input layer.
• Hidden layers: Hidden layers process and transport data to other layers.
• Output layer: The result or prediction is made in the output layer.
Examples

• Self-Driving Cars
• Face Recognition
• Chatbots
RNN (Recurrent Neural Network) is a neural network
where output is predicted based on precious outputs. It is
useful for tasks like predicting the next word in a
sentence.

CNN (Convolutional Neural Network) are neural


networks used to identify patterns in data and used for
DEEP feature extraction.
LEARNING
ALGORITHMS LSTM (Long Short-Term Memory) is a type of RNN.I t
can remember information for longer periods.

CNN-BiLSTM (Convolutional Neural Network -


Bidirectional Long Short-Term Memory) was used to
improve performance to detect patterns in both states
(forward and reverse direction)
Previous Research Work
In research 14 NSL-KDD datasets were used and different algorithms such as Vanilla RNN, SVM, DNN and GRU-RNN
Research were employed. In that, GRU-RNN achieved the highest accuracy of 89%.

Research Research 15 implemented two deep learning models in CICDDo2019 dataset obtained accuracy of 99 percent. •

Research 16 has experimented with just LSTM in the CICDS2017 dataset. They tried different number of hyper
Research parameters and LSTM layers.

Research 17 presents a deep learning algorithm in which they merged RNN with CNN. This combination outdid the
Research other models.

In Research 18 the author used a set of deep and machine learning algorithms to detect anomalies in IOT devices using
Research CICIDS2017 dataset.
• The previous research work in detecting
DDOS attacks has proved that deep learning
GAP IN algorithms are better than traditional
PREVIOUS machine learning algorithms to detect
DDOS attacks and mitigate them, but they
RESEARCH cannot detect DDOS attacks without
WORK affecting legitimate traffic.
In this research, three deep learning
algorithms CNN, RNN and LSTM are used to
make CNNBiLSTM model. The main purpose
of this model was to detect DDOS attacks on
Work the network without affecting legitimate
traffic.
Incoming data has been verified by applying
Done In different processing steps on it such as
checking and filtering of data, Nan and
this infinite values were replaced, labels were
encoded, and the data was split into train -test
Research sets.
Data was reshaped so that it can be fit into
each model. Algorithms RNN, CNN and
LSTM, CNN- BiLSTM were used to make
one standard model.
The CNN- BiLSTM model showed the best
performance, followed by LSTM and RNN,
outperforming most other methods.

Advantages This Model can detect DDOS attacks on a


network without affecting legitimate traffic
and
Research The main limitation of the proposed model is
the lack of a realistic testing platform, which
Gap affects the reliability of the testing results.

IOT networks are still vulnerable and require


resistance abilities against cyber-attacks.
Thank you!

You might also like