Abstract:Abstract With the rapid growth of informatics systems’ technology in this modern age, the Internet of Things (IoT) has become more valuable and vital to everyday life in many ways. IoT applications are now more popular than they used to be due to the availability of many gadgets that work as IoT enablers, including smartwatches, smartphones, security cameras, and smart sensors. However, the insecure nature of IoT devices has led to several difficulties, one of which is distributed denial-of-service (DDoS) attacks. IoT systems have several security limitations due to their disreputability characteristics, like dynamic communication between IoT devices. The dynamic communications resulted from the limited resources of these devices, such as their data storage and processing units. Recently, many attempts have been made to develop intelligent models to protect IoT networks against DDoS attacks. The main ongoing research issue is developing a model capable of protecting the network from DDoS attacks that is sensitive to various classes of DDoS and can recognize legitimate traffic to avoid false alarms. Subsequently, this study proposes combining three deep learning algorithms, namely recurrent neural network (RNN), long short-term memory (LSTM)-RNN, and convolutional neural network (CNN), to build a bidirectional CNN-BiLSTM DDoS detection model. The RNN, CNN, LSTM, and CNN-BiLSTM are implemented and tested to determine the most effective model against DDoS attacks that can accurately detect and distinguish DDoS from legitimate traffic. The intrusion detection evaluation dataset (CICIDS2017) is used to provide more realistic detection. The CICIDS2017 dataset includes benign and up-to-date examples of typical attacks, closely matching real-world data of Packet Capture. The four models are tested and assessed using Confusion Metrix against four commonly used criteria: accuracy, precision, recall, and F -measure. The performance of the models is quite effective as they obtain an accuracy rate of around 99.00%, except for the CNN model, which achieves an accuracy of 98.82%. The CNN-BiLSTM achieves the best accuracy of 99.76% and precision of 98.90%.

DeepAK-IoT: an Effective Deep Learning Model for Cyberattack Detection in IoT Networks

An efficient deep recurrent neural network for detection of cyberattacks in realistic IoT environment

Evaluating deep learning variants for cyber-attacks detection and multi-class classification in IoT networks

IoT-Based Intrusion Detection System Using New Hybrid Deep Learning Algorithm

An Efficient Deep Learning Approach to IoT Intrusion Detection

Security-aware IoT botnet attack detection framework using dilated and cascaded deep learning mechanism with conditional adversarial autoencoder-based features

A hybrid deep learning-based intrusion detection system for IoT networks

Performance Evaluation of Deep Learning Techniques in The Detection of IOT Malware

Enhancing Cybersecurity in IoT Networks: A Deep Learning Approach to Anomaly Detection

Intrusion Detection in IoT Using Deep Residual Networks with Attention Mechanisms

Machine learning-based attack detection for the Internet of Things

Deep learning in distributed denial-of-service attacks detection method for Internet of Things networks

Deep Learning based DDoS Attack Detection in Internet of Things: An Optimized CNN-BiLSTM Architecture with Transfer Learning and Regularization Techniques

AI-Based Intrusion Detection for a Secure Internet of Things (IoT)

Deep Learning-Inspired IoT-IDS Mechanism for Edge Computing Environments

Deep learning approach for intrusion detection in IoT-multi cloud environment

Deep learning model for elevating internet of things intrusion detection

Approach for Detecting Attacks on IoT Networks Based on Ensemble Feature Selection and Deep Learning Models

A State-of-the-Art Technique for Malware Detection Based on Deep Learning in IoT Environment

Deep Learning in IoT Intrusion Detection

A Novel Deep Learning-Based Intrusion Detection System for IoT Networks