Internet of things (IoT) is penetrating into every aspect of our lives including our body, our home and our living environment along with numerous security challenges. With rapidly growing number of connected devices in IoT, the scope for cyber-attack also increases exponentially. Therefore an effective intrusion detection system (IDS) is needed to efficiently detect the attack at faster rate in highly scalable and dynamic IoT environment. In this paper, a novel intrusion detection technique is proposed based on fog computing using Online Sequential Extreme Learning Machine (OS-ELM) which can intelligently interpret the attacks from the IoT traffic. In the proposed system, the existing centralized cloud intelligence in detecting the attack is distributed to local fog nodes to detect the attack at faster rate for IoT application. The distributed architecture of fog computing enables distributed intrusion detection mechanism with scalability, flexibility and interoperability. The analysis of the proposed system proves to be efficient in terms of response time and detection accuracy.
Our world is moving towards an Internet of Things (IoT) era by connecting billions of IoT. There are several security loopholes in the IoT network. Intrusion can lead to performance degradation and pose a threat to data security. Hence, there is a need for a method to detect intrusion in the IoT networks. Existing solutions use supervised learning based intrusion detection methods that need a huge labeled dataset for better accuracy. It is not easy to source out a huge labeled dataset because the size of the IoT network is huge. To overcome some of the impediments in the existing solutions, we propose a novel SDRK Machine Learning (ML) algorithm to detect intrusion. SDRK leverages supervised deep neural networks and unsupervised clustering techniques. The intrusion detection and mitigation algorithms are placed in the fog nodes that are between IoT and cloud layers. We test our proposed methodology against the Data Deluge (DD) attack in the testbed. SDRK model is tested on the benchmark NSL-KDD dataset. We compare the results with state-of-the-art solutions. When testing with the NSL-KDD dataset, we find that SDRK detects the attacks with improved accuracy of 99.78%.
Information security deals with a large number of subjects like spoofed message detection, audio processing, video surveillance and cyber-attack detections. However the biggest threat for the homeland security is cyber-attacks. Distributed Denial of Service attack is one among them. Interconnected systems such as database server, web server, cloud computing servers etc., are now under threads from network attackers. Denial of service is common attack in the internet which causes problem for both the user and the service providers. Distributed attack sources can be used to enlarge the attack in case of Distributed Denial of Service so that the effect of the attack will be high. Distributed Denial of Service attacks aims at exhausting the communication and computational power of the network by flooding the packets through the network and making malicious traffic in the network. In order to be an effective service the DDoS attack must be detected and mitigated quickly before the legitimate user access the attacker's target. The group of systems that is used to perform the DoS attack is known as the botnets. This paper introduces the overview of the state of art in DDoS attack detection strategies.
26 May 2020 in IEEE Transactions on Network and Service Management
Link Flooding Attack (LFA) is a genre of Distributed Denial of Service (DDoS) attack. LFA can cut off a target area from the network, without directly attacking the target. The attacker chooses links which when cut off will disconnect the target area and instruct the bots to flood those links with small packets. Some of the existing solutions are suitable for specific routing methods like shortest path routing or need cooperation between Autonomous Systems (AS). To overcome certain hitches of existing solutions, we have proposed a novel mechanism named BALANCE. It detects and mitigates LFA via hybrid-Software-Defined Network (SDN). SDN splits the control and data plane using OpenFlow protocol. Hybrid SDN has both legacy and SDN nodes, with a controller in the control plane. We have used Service Based Hybrid SDN (SBHS), which is a type of hybrid-SDN. BALANCE begins with an algorithm that chooses nodes in an AS to be SBHS enabled in such a way that the controller can get statistics of all the links in the AS. Next, congestion detection and location algorithms are implemented in the controller to find the congested links. Finally, LFA bot detection and mitigation algorithms are implemented in the controller to mitigate LFA. BALANCE was evaluated in testbed and emulator. We compared the results with state-of-the-art solutions. BALANCE was able to detect LFA bots at a precision of 97.64% and had HTTP response time of 2 seconds during the LFA attack.
10 November 2020 in IEEE Transactions on Network and Service Management
Software-Defined Network (SDN) segregates the control plane and the data plane to bring about a programmable network. The controller at the control plane runs network modules and sets rules for forwarding the packets in the switches that resides at the data plane. Though advantageous in several ways, SDN can fail when the controller is saturated by a flood of TCP SYN packets. SYN flood can be created using malicious spoofing of IP or MAC addresses or flash crowd. The existing solutions to mitigate SYN flood against the controller does not adequately handle MAC spoofing based SYN flood, and these are unable to distinguish between flash crowd and malicious traffic. To overcome some limitations in existing solutions, we propose a novel mechanism called AEGIS, which detect and mitigate SYN flood against the controller in SDN. AEGIS runs in the controller, and it regularly checks if there is a performance lag in the controller due to an ongoing SYN flood. If a performance degradation is detected, then AEGIS takes it an indication of SYN flood and it identifies whether it is due to spoofed addresses or flash crowd. Once the reason is found, the appropriate mitigation procedure is triggered. We evaluate AEGIS in testbed and emulator settings, and we compare the results of the evaluation with state-of-the-art solutions. The performance evaluation of AEGIS shows that it identifies the malicious SYN at an accuracy of 97.78%. Moreover, when there is no SYN flood, AEGIS takes 0.0637s to set up a successful TCP connection, which is 53.81% less than the time taken by the state-of-the-art solution, thus, it proves that AEGIS is lightweight.