Multiclass Classification Procedure for Detecting Attacks on MQTT‐IoT Protocol

Aláiz‐Moretón, Héctor; Aveleira‐Mata, José; Ondicol-Garcia, Jorge; Muñoz-Castañeda, Angel Luis; García-Rodríguez, Isaías; Benavides, Carmen

doi:10.1155/2019/6516253

Cited by 96 publications

(62 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In comparising with simular works, our approach demonstrates the better results in DoS attack detection for some indicators (optimal time interval and classification method). For instance, in our research F1-score value reached the value of 0.98 whereas the results of [27] shows only 0.90 and research results of [28] being equal to 0.95.…”

Section: Results and Analysiscontrasting

confidence: 58%

DoS attacks detection in MQTT networks

Dikii

Arustamov

Grishentsev

2021

IJEECS

View full text Add to dashboard Cite

<span>The paper considers the problem of protecting the Internet of things infrastructure against denial-of-service (DoS) attacks at the application level. The authors considered parameters that affect the network gateway workload: message frequency, payload size, number of recipients and some others. We proposed a modular structure of the attack detection tool presented by three classifiers that use the following attributes: username, device ID, and IP-address. The following types of classifiers have been the objects for the research: multilayer perceptron, random forest algorithm, and modifications of the support vector machine. Some scenarios for the behavior of network devices have been simulated. It was proved that for the proposed feature vector on simulated training and test data sets, the best results have been shown by a multilayer perceptron and a support vector machine with a radial basis function of the kernel and optimization with SMO algorithm. The authors also determined the conditions under which the selected classifiers have the best quality of recognizing abnormal and legitimate traffic in MQTT networks. </span>

show abstract

Section: Results and Analysiscontrasting

confidence: 58%

DoS attacks detection in MQTT networks

Dikii

Arustamov

Grishentsev

2021

IJEECS

View full text Add to dashboard Cite

show abstract

“…the best combinations of the param-grid found, will have the minor MAE, obtaining the best model. The hyper-parameter set of the three implemented machine learning techniques are described at [56], [57], [58] and [59].…”

Section: Methodsmentioning

confidence: 99%

Hydrogen consumption prediction of a fuel cell based system with a hybrid intelligent approach

Sousa

Aláiz‐Moretón

Quintián

et al. 2020

Energy

View full text Add to dashboard Cite

Energy storage is one of the challenges of the electric sector. There are several different technologies available for facing it, from the traditional ones to the most advanced. With the current trend, it is mandatory to develop new energy storage systems that allow optimal efficiency, something that does not happen with traditional ones. Another feature that new systems must meet is to envisage the behavior of energy generation and consumption. With this aim, the present research deals the hydrogen consumption prediction of a fuel cell based system thanks a hybrid intelligent approach implementation.

show abstract

“…Then an update algorithm will be applied to the weights to decrease the loss and eventually increase the accuracy. Categorical cross-entropy and binary crossentropy are used in the DLS-IDS model to evaluate the loss in the case of multiclass and binary classification, respectively, as follows [33]: (17) where ( ,̂) is the loss function, , ̂ represents the actual and predicted output of sample for class , respectively. The binary cross-entropy is given by…”

Section: ) Loss Computationmentioning

confidence: 99%

Implementing a Deep Learning Model for Intrusion Detection on Apache Spark Platform

2020

View full text Add to dashboard Cite

Internet evolution produced a connected world with a massive amount of data. This connectivity advantage came with the price of more complex and advanced attacks. Intrusion Detection System (IDS) is an essential component for security in modern networks. The IDS methodology is either signature-based detection or anomaly behavior detection. Recently, researchers adopted Deep Learning (DL) because it has a better performance than traditional machine learning algorithms. The use of DL to produce a model for the IDS may take a long time because of computation complexity and a large number of hyperparameters. Different DL models for IDS on Apache Spark have been implemented in this paper. This paper uses the famous Network Security Lab-Knowledge Discovery and Data Mining (NSL-KDD) dataset and presents a computation delay comparison between Apache Spark and regular implementation. Moreover, an enhanced model is used to improve attack detection accuracy.

show abstract

Multiclass Classification Procedure for Detecting Attacks on MQTT‐IoT Protocol

Cited by 96 publications

References 42 publications

DoS attacks detection in MQTT networks

DoS attacks detection in MQTT networks

Hydrogen consumption prediction of a fuel cell based system with a hybrid intelligent approach

Implementing a Deep Learning Model for Intrusion Detection on Apache Spark Platform

Contact Info

Product

Resources

About