What are DDoS attacks and Software Defined Networks(SDN)?

1. INTRODUCTION
a. CONTEXT

The concept of “software-defined networks” (SDNs) has recently attracted a lot of attention in the academic community.

When using the SDN paradigm, it is much simpler to design a resilient, adaptable, and safe network infrastructure. The key innovation of SDN is the decoupling of the control plane from the underlying infrastructure. The final approval for publishing was given by assistant editor Luis Javier Garcia Villalva. The machine learning (ML) technique simplifies the development of network behavior with the ability to learn from data history and produce packet-level predictions based on training data.

DDoS attacks, which disrupt service for many users at once, have become a focal point of cyber security discussion in recent years. Software-defined networking (SDN) has been the subject of much study and development in recent years. Taking advantage of the separation of the control and data planes in an SDN, Shin and Gu developed a network scanning tool capable of identifying such networks and then conducted a denial-of-service attack against one. To find methods for DDoS detection in an SDN network, the authors provided a variety of characteristics and analyzed traffic behavior during a DDoS assault. Researchers have also suggested DDoS detection systems using severity of attack (DDADA) and machine learning (called DAMDL).

DDoS assaults are becoming sophisticated and widespread, making them a major problem for modern enterprises. In order to rapidly and correctly identify DDoS assaults in SDNs, the authors of this piece propose a deep CNN architecture. Additionally, a technique is being developed to detect flow-based DDoS assaults using a deep CNN ensemble. Similar state-of-the-art concatenations of DL-based ensembles and hybrid algorithms (such as RNN, LSTM, and RL) are employed for verification. One of the most difficult types of distributed denial of service (DDoS) attacks to detect is the low-rate kind (LR-DDoS). In contrast to high-rate DDoS assaults, low-rate DDoS (LR-DDoS) attacks do not overwhelm a network with an overwhelming volume of data. Instead, it selectively activates parts of the protocol, such as TCP’s retransmission after timeout and congestion control.

By physically separating the control and forwarding planes of network devices, the relatively new networking concept known as software-defined networking (SDN) intends to create a logically centralized control and management entity (such as routers and switches). Meanwhile, SDN might be put to use in the detection and mitigation of DDoS assaults that make limited use of network resources.

b. PROBLEM

Because the SDN network is different from the traditional network in how it is built, DDoS attacks could make it hard for the SDN to work. The SDN controller is the part that is most likely to be attacked by a DDoS. In general, a DoS attack tries to block real users from using network resources. In order to launch a distributed denial of service assault, attackers often use a large number of dump terminals, computers, and botnets all at once. This consumes the primary resources of the system and causes the cessation of all services.

DDoS assaults, both small and large, may be carried out using a variety of legitimate and efficient methods. The perpetrators of a recent distributed denial-of-service assault employed legitimate Memcached technology. Displays the recommended ensembles, including both pure and hybrid ensembles, whose primary objective is to lessen demands on the network’s underlying resources. Answers were delivered to the intended addresses at a rate of 126.9 million packets per second, thanks to the attacker’s use of Memcached objects and IP address spoofing. This depleted the victim’s energy and time significantly. Since DDoS attacks often employ spoofed IP addresses, they are difficult to trace. This makes detecting, preventing, and avoiding DDoS assaults difficult and time-consuming. Strong, creative methods of detection, defenses that may halt or slow down, and mitigation strategies can all be used to swiftly locate complicated DDoS assaults.

A DDoS assault may be detected in one of two ways: either by signatures or abnormalities. The first method compares incoming data against signatures consisting of patterns or strings extracted from protocol header fields in order to identify malicious flows (or not). Low-Rate Distributed Denial-of-Service Attack Detection and Mitigation Architecture One of the most challenging aspects of employing ML algorithms to detect LR-DDoS is selecting appropriate threshold values, since these methods rely on them. Coming up with realistic solutions to LR-DDoS assaults is similarly challenging. Although updating the router’s firmware is sometimes feasible, it is not always a viable option.

c: OBJECTIVES
  • This research will examine four factors to determine the efficacy of DDoS attack detection while the SDN controller is under assault (named flow length, flow duration, flow size, and flow ratio). To aid in identifying a Distributed Denial of Service (DDoS) assault, the concept of “degree of attack” is proposed.
  • This research will provide a means of learning about the assault, the severity of which should be taken into account (dubbed DDADA). Finally, a novel machine learning-based detection approach called DDAML is utilized to locate the DDoS assault, enhancing the effectiveness of detection.
  • A cutting-edge system that employs deep learning to detect SDN DDoS assaults This system will employ novel ensemble CNN models for improved flow-based data identification.
  • A look at how the suggested solution stacks up against state-of-the-art deep ensembles and hybrid DDoS attack detection systems in SDNs Ensemble architecture based on the suggested SDN controller must be scalable and cost-effective (i.e., control plane).
METHODOLOGY

Four articles were analyzed by researchers. A secondary dataset of four articles was used for this report. Several methods for detecting distributed denial of service attacks were developed by the research community. Examination of these techniques revealed that while behavioral aspects were crucial for DDoS detection in SDN, they were also influenced by other parameters. In order to discover methods for detecting DDoS assaults in the SDN network, researchers in this study included many attributes and analyzed traffic patterns during an attack. In addition, a machine learning-based DDoS detection algorithm and a DDoS detection algorithm based on the degree of attack (abbreviated DDADA) have been proposed (called DAMDL). The proposed techniques allow one to detect DDoS assaults in an SDN setting.

In this investigation, a modular and malleable security architecture was developed and deployed to detect and thwart LR-DDoS assaults in SDN environments. The modular structure allows for easy customization without disrupting the overall architecture. In order to identify flows, the architecture’s IDS component employs a collection of trained ML models that may be constructed using a variety of languages and frameworks.

A method for preventing SDN-based, low-rate DDoS assaults has been developed by researchers. The framework decouples issue detection and resolution in network applications and may be implemented in any programming language or technology. The controller’s processing load is therefore reduced.

3. ANALYSIS

According to the experimental findings, the proposed model outperforms single-SVM in terms of classification accuracy and general effectiveness. Future-proof security policies that include the proposed model for the controller are possible. The training and testing data came from a contemporary DDoS dataset with 1,216,666 records and 27 attributes. As an example, NSL-KDD encompasses a wide variety of attacks, including Probe, DoS, R2L, and U2R. Since there are no duplicates in either set of data, the machine learning classifiers will be less likely to overweight frequent occurrences. As a result, it’s possible that many ML algorithms will function as intended. An accurate machine learning (ML) model may predict the nature of an impending assault.

The evaluation found that even though LR-DoS assaults are difficult to uncover, our technique had a 95% detection rate. According to the study authors, the open network operating system (ONOS) controller installed on the Mininet virtual machine helps make the deployment as realistic as feasible. In our testing topology, the intrusion prevention detection system handles any threats previously identified by the IDS. This demonstrates that our system can effectively detect and halt LR-DDoS assaults.

The adoption of hybrid approaches has increased the precision of nanomomals to 98.75%. In comparison to RNN ensembles, LSTM ensembles, RL ensembles, and hybrid RL ensembles, our recommended CNN ensemble has a detection accuracy of 99.45 percent.

4. CONCLUSION

The DDoS attack is now the biggest threat to the SDN network’s security. Even the most advanced methods for spotting DDoS attacks have low accuracy and can be thrown off by other things. We propose a suite of deep learning ensemble-based detection and prevention strategies for scaling large-scale networks. In contrast to conventional wisdom, the proposed method actually makes it more difficult to deduce the necessary steps. The modular structure allows for easy customization without disrupting the overall architecture.

In order to identify flows, the architecture’s IDS component employs a collection of trained ML models. Six alternative ML algorithms were evaluated on the CIC DDoS dataset, with findings showing that 95% of them were accurate. Modern machine learning and deep learning techniques are something that researchers want to include in their work in the near future.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top