Types of Attacks against Federated Neural Networks and Protection Methods

The article discusses key issues of web application security and methods of protection against hacker attacks. With businesses increasingly dependent on online services, web application security is becoming one of the top priorities for organizations. Sufficient attention should be paid to security measures to protect user data from unwanted access and malicious use. This article reviewed the main types of hacker attacks on web applications, such as SQL injection (SQLi), Cross-site scripting (XSS), Cross-site request forgery (CSRF), Server-side request forgery (SSRF) and file attacks, as well as examples of protection measures against each type of attacks. Updated and improved basic principles of web application security are proposed based on threat modeling techniques.

Phishing, as a type of information attack, has been used by intruders for selfish purposes for quite some time. They are very popular in the criminal world because it is much easier for a person to make certain profitable actions than a program. With the advent of new technologies, this type of attack has gradually adapted to the new conditions of engagement with its victim. Cloud services have become a great modern and widespread tool for phishing campaigns. The use of such services has given to malicious actors a number of significant advantages over the use of their own computing resources. The relative cheapness and ease of exploitation of these technologies has played an important role. The problem of information security with using cloud technologies is that this type of attack is difficult to detect, even more to prevent, without significantly affecting the comfort of using end users of information systems. The article analyzes the relevance of this type of attacks based on real data. We considered the algorithm of their work during a life cycle and analyzes the use of the basic available security methods of protection, their feasibility and problems of use. The analysis showed that not all modern security methods are capable of detecting and preventing phishing attacks, which use public cloud services. Even a combination of several or all methods cannot guarantee high protection for users against phishing threats. In the article were mentioned some examples of phishing campaigns that took place during 2019 and used such popular public cloud services as Azure Blob storage created by Microsoft and Google Drive developed by Google. A basic list of tips was also provided that would increase the level of security for internet users in order to reduce the risk of potential data compromise or its consequences.

The article analyzes the main types of social engineering attacks and their classification. An overview of the instruments for detecting and counteracting social engineering as a method of accessing confidential information is provided, and key tools for counteracting such attacks are proposed. Experiments are conducted to demonstrate the effectiveness of each tool in different types of attacks. Particular attention is paid to software tools that help minimize the risks and losses from social engineering attacks. A system of integrated use of security tools has been developed that provides almost complete protection of the information system from such threats. The use of the WAZUH SIEM system in combination with the VirusTotal module provides opportunities to detect and counteract most types of attacks, including social engineering. For its part, integrating the Mozilla Firefox browser with the Startpage Privacy Protection application guarantees anonymous web browsing. The effectiveness of the protection methods has been confirmed experimentally. The proposed system was tested by modeling attacks, which proved its effectiveness. It was found that the integrated use of all modules provides the maximum level of protection of information resources. At the same time, the absence of any module reduces the overall security level.

The recent rapid development of cloud technologies has encouraged its widespread use by individual mobile users, private organizations and public institutions. Mobile users and organizations deploy their data on cloud servers and use it. Connections to cloud servers are realized over the Internet, which makes data transmitted over the network vulnerable to various types of attacks. Although numerous security solutions have been proposed for data security in cloud computing systems, the security of provided services remains an actual problem for both cloud users and cloud service providers. The article provides a general survey of security and privacy issues in cloud computing systems, and reviews various types of attacks and possible threats, as well as protection methods and available solutions against such attacks, and proposes mechanisms.

The purpose of the study: development and evaluation of a method for countering FGSM, ZOO, OPA adversarial attacks on image classification systems based on the integration of noise pollution, neural cleansing and JPEG data compression. Research methods: system analysis, machine learning, image noising, neural cleansing, JPEG data compression, computational experiment. Results obtained: an analysis of works on the topic of attacks on image classification systems (ICS) based on the application of machine learning methods, and methods of protection against them was carried out. Based on the results of this analysis, it was revealed that the most common attacks on ICS include adversarial attacks, namely: Fast Gradient Sign Method (FGSM), Zero-Order Optimization (ZOO) and One Pixel Attack (OPA). The topic of countering these attacks is currently of great interest. The essence of the impact of these attacks on ICS is disclosed, and their influence on the accuracy of ima ge recognition is revealed. A method for countering adversarial attacks is proposed, based on image noising with Gaussian and Poisson noise, as well as the use of JPEG compression and neural cleansing technology. Experiments were conducted showing the high efficiency of the proposed method. The experiments were aimed at assessing the accuracy of image re cognition contained in two different data sets - a set of images of personal computer parts and a set of handwritten digital images. The results of image recognition were evaluated before and after exposure of the ICS to adversarial attacks, as well as after applying the proposed method to these sets. Scientific novelty: an analysis of works on the topic of protection against adversarial attacks showed that currently the most typical attacks on ICS are FGSM, ZOO and OPA attacks. The proposed method for countering adversarial attacks on ICS differs from other known protection methods in that it integrates the capabilities of countering attacks contained in three different approaches (neural cleansing, noise pollution and JPEG compression) and identifies the optimal para meters of these approaches. The high efficiency of the proposed method was confirmed in experiments conducted on two different data sets. Contribution: Igor Kotenko and Igor Saenko – general concept of adversarial attacks on ICS and methods of protection against them based on well-known works; Igor Kotenko and Oleg Lauta – description of methods of impact of adversarial attacks; Nikita Vasilev and Vladimir Sadovnikov – implementation of the proposed approach; Igor Kotenko and Igor Saenko – theoretical justification of the proposed approach.

With the extensive application of Internet of Things and wireless sensor networks (WSNs) in various real-time fields, they have become increasingly vulnerable to several types of attacks that could have a serious impact in their functionalities. Recently, intrusion detection systems have become one of the crucial security components. We propose in this work an approach based on machine and deep learning for DoS and DDoS attack detection. We have evaluated, analyzed, and compared the efficiency of three learning models separately, including deep neural network (DNN), random forest, and decision tree, using the standard metrics of evaluation such as accuracy, precision, F1-score, and recall. In our contribution, an approach ensemble learning was introduced in order to enhance the accuracy rate of classification. Our study was carried out using two well-known benchmark and real-time datasets, CICIoT-2023 and WSN-DS, intended for wireless sensor networks and IoT. These datasets containing various types of attacks. CICIoT2023 dataset contains 33 attacks divided into 7 classes, while WSN-DS dataset contains four types of DoS attacks, including Blackhole, Grayhole, Flooding, and TDMA. The experiment result demonstrate the effectiveness of our approach in attacks detection with high accuracy achieved close to perfect. Our approach result demonstrates that using the ensemble learning technique works better than each ML architecture independently.

A denial-of-service attack (DoS attack) or distributed denial- of-service attack (DDoS attack) is an attempt to make a computer resource unavailable to its intended users [1]. There are many types of Denial of Service Attacks. [2] gives a very detailed explanation of different types of attacks. We focus mainly on the spoofing related attacks and our observation on the pattern of the different types of DoS attacks shows that with minor changes in the program we could generate all the other types of attacks. Here are few DoS attacks that we would like to mention whose signature and the procedure of generating is described at [2].

Internet of Things (IoT) is one of the fastest-growing technologies. With the deployment of massive and faster mobile networks, almost every daily-use item is connected to the Internet. IoT-enabled industrial multimedia environment is used for the collection and analysis of different types of multimedia data (i.e., images, videos, audios, etc.). This multimedia data is generated by various types of smart devices like drones, robots, smart controller, smart surveillance system which are deployed for the industrial monitoring and control. The multimedia data is generated in the enormous amount which can be considered as the big data. This data is further utilized in various types of business needs for example, chances of fire accidents in the industrial plant, overall machine health, etc., which can be predicted through the application of big data analytics. Therefore, IoT-enabled industrial multimedia environment is very helpful to the concerned authorities as they come to know the important information in advance. However, all the smart devices are connected and controlled through the Internet. It further causes severe threats to the communication happens in an IoT-enabled industrial multimedia environment. It is vulnerable to various types of attacks such as replay, man-in-the-middle, impersonation, secret information leakage, sensitive information modification, and malware injection (i.e., mirai). Therefore, it is important to prevent the communication of such an environment against the different types of possible attacks. These days, the attacks performed by botnets (i.e., malware attacks such as mirai and reaper) have drawn attention to the researchers. Under the influence of such attacks, the communication of IoT-enabled industrial multimedia environment is disrupted. Moreover, the attackers may also control the smart devices remotely and can change their functionalities. Hence, we need some robust mechanism to detect the presence of the malware attacks in such an environment. In this paper, we propose a malware detection mechanism in IoT-enabled industrial multimedia environment with the help of machine-learning approach, which is named as MADP-IIME. MADP-IIME uses four different types of machine learning methods (i.e., naive bayes, logistic regression, artificial neural networks (ANN) and random forest) to detect the presence of malware attacks successfully. Furthermore, MADP-IIME performs better than other related existing schemes and achieves $$99.5 \%$$ detection and $$0.5 \%$$ false positive rate. In addition, the conducted security analysis proves the resilience of the proposed MADP-IIME against different types of malware attacks.

With the increasing use of deep learning techniques in real-world applications, their vulnerabilities have received significant attention from deep-learning researchers and practitioners. In particular, adversarial examples for deep neural networks and protection methods against them have been well-studied in recent years because they have serious vulnerabilities that threaten safety in the real-world. Audio adversarial examples, which are targeted attacks, are designed such that the deep neural network-based speech-to-text systems misunderstand input voice sound. In this study, we propose a new protection method against audio adversarial examples. The proposed protection method is based on a sandbox approach, where an input voice sound is checked in the system to determine if it is an audio adversarial example. To evaluate the proposed protection method, we used actual audio adversarial examples created on deep speech, which is a typical speech-to-text transcription neural network. The experimental results show that our protection method can detect audio adversarial examples with high accuracy.

Wi-Fi networks are often susceptible to signal interference that can disrupt services, intercept sensitive data, and allow unauthorized access. Exploiting this vulnerability, attackers can launch Intentional Electromagnetic Interference (IEMI) attacks or protocol-based attacks. IEMI attacks include, but are not limited to, jamming, which involves the introduction of interference signals in the frequency band used by a given Wi-Fi network, thus disrupting its proper operation. The article presents research on different types of IEMI attacks, aimed at illustrating this type of threat to video transmission from Wi-Fi cameras and discusses methods of protection against this type of attack. The paper examines the methods of jamming the wireless network band, which are used on the distribution methods and the Sweeping Jammer methods that cyclically disrupt all available devices. Based on experiments, it has been confirmed that SOHO devices are highly susceptible to this type of simple attacks and implementing a method to minimize the effects of jamming yields poor results. The simplicity of this type of attack and the high availability of software and hardware tools on the market make the threat highly popular and still dangerous.

Web resources are an integral part of the life of a modern person, who are now more and more often subjected to hacker attacks. The most popular types of attacks are the SQL-injections and cross-site scripting, but DDoS attacks continue to be in the top 10 of network attacks and lead to serious crashes and failures of web resources. The most common type of DDoS attack is UDP flood attack, based on the infinite sending of UDP packets to ports of various UDP services. The scientific novelty of the work lies in the fact that to increase the level of security of web resources a medium-term forecast of cyber attacks of the UDP-flood type, using the methods of correlation analysis, based on the additive time series model, is proposed taking into account seasonal factors and time duration, which will ensure the necessary level of web security -resources. The type of UDP-flood attacks was chosen as the object of study. Using the methods of correlation analysis and modeling, we calculated the seasonal index of UDP flood attacks, and the autocorrelation of the time series of this type of attack. Using the method of simple exponential smoothing, a forecast of UDP-flood attacks is constructed. The paper proposes a classification of DDoS attacks, describes protection methods. Based on the correlation analysis, the predicted values of the impact of UDP flood attacks on web resources are calculated, and the seasonal factor is revealed. The largest number of attacks is expected in the IV quarter of 2020. For DDoS attacks lasting up to 20 minutes, seasonality was also revealed in the 1st quarter of the calendar year, which means that in the 1st quarter of 2020 the largest number of attacks of this duration should be expected. Prospects for further research into the problem of protection against DDoS attacks are presented in the further development of the methodology for countering UDP flood attacks and information security algorithms for web resources, which will reduce the number of UDP flood attacks and increase the level of web resource security.

Federated learning is a technology for privacy-preserving learning in distributed storage systems. This training allows you to create a general forecasting model, storing all the data in your storage systems. Several devices take part in training the general model, and each device has its own unique data on which the neural network is trained. The interaction of devices occurs only to adjust the weights of the general model. After which, the updated model is transmitted to all devices. Training on multiple devices creates many attack opportunities against this type of network. After training on a local device, model data is sent via some type of communication to a central server or global model. Therefore, vulnerabilities in a federated network are possible not only at the training stage on a separate device, but also at the data exchange stage. All this together increases the number of possible vulnerabilities of federated neural networks. As is known, not only neural networks, but also other models can be used to build federated classifiers. Therefore, the types of attacks directly on the network also depend on the type of model used. Federated neural networks are a rather complex design, different from neural networks and other classifiers, which can be vulnerable to various types of attacks because training occurs on different devices, and both neural networks and simpler algorithms can be used. In addition, it is necessary to ensure data transfer between devices. All attacks come down to several main types that exploit classifier vulnerabilities. It is possible to implement protection against attacks by improving the architecture of the classifier itself and paying attention to data encryption.

Adversarial learning is a recently introduced term which refers to the machine learning process in the presence of an adversary whose main goal is to cause dysfunction to the learning machine. The key problem in adversarial learning is to determine when and how an adversary will launch its attacks. It is important to equip the deployed machine learning system with an appropriate defence strategy so that it can still perform adequately in an adversarial learning environment. In this paper we investigate artificial neural networks as the machine learning algorithm to operate in such an environment, owing to their ability to learn a complex and nonlinear function even with little prior knowledge about the underlying true function. Two types of adversarial attacks are investigated: targeted attacks, which are aimed at a specific group of instances, and random attacks, which are aimed at arbitrary instances. We hypothesise that a neural ensemble performs better than a single neural network in adversarial learning. We test this hypothesis using simulated adversarial attacks, based on artificial, UCI and spam data sets. The results demonstrate that an ensemble of neural networks trained on attacked data is more robust against both types of attack than a single network. While many papers have demonstrated that an ensemble of neural networks is more robust against noise than a single network, the significance of the current work lies in the fact that targeted attacks are not white noise.

The article examines the challenges of global navigation satellite systems (GNSS) functioning at sea under unintentional and intentional interference. The article reviews the vulnerability of GNSS, the methods of protection against interference and the ways of mitigating their impact based on the marine concept of positioning, navigation and time synchronization (PNT). The main goal of this concept is the guaranteed obtaining of reliable data on coordinates, navigation and exact time due to the combined use and comparison of indication for disparate systems and sensors under the influence of natural or intentional interference (attacks) on ship's GNSS equipment. The article analyzes various open sources of information to identify two methods of ensuring the integrity and accuracy of PNT data according to IMO documents and standards. The first method is the detection and direct countermeasures against attacks on shipboard GNSS equipment. The article determines that the most common and easy-to-implement type of attacks are jamming attacks of satellite signals, unlike the more complex and challenging spoofing attacks. The main approach to protection against jamming attacks is spatial signal processing using adaptive marine antenna arrays with a controlled pattern. Examples of modern practical developments of adaptive antenna arrays are given. The second method to ensure reliable PNT data is the use of alternative navigation systems, redundant capabilities and non-traditional methods using the existing systems. Technical solutions in this method have limitations due to the requirements for the vessel’s conventional navigation and radio communication installation. IMO has suggested structures of multisensor and multisystem receivers for obtaining reliable PNT data. These structures combine primary data from different systems based on different principles, such as satellite, terrestrial and augmented correction systems, vessel navigation data and reference systems. The processed PNT data must be accompanied by accuracy and integrity indicators. Keywords: positioning, navigation, time synchronization, jamming, spoofing, spatial processing, antenna arrays, cyber risks, information protection.

Advances in technology have led not only to increased security and privacy but also to new channels of information leakage. New leak channels have resulted in the emergence of increased relevance of various types of attacks. One such attacks are Side-Channel Attacks, i.e. attacks aimed to find vulnerabilities in the practical component of the algorithm. However, with the development of these types of attacks, methods of protection against them have also appeared. One of such methods is White-Box Cryptography.