Malicious User Attack Research Articles

Blockchain-Inspired Trust Management in Cognitive Radio Networks with Cooperative Spectrum Sensing

Cooperative spectrum sensing (CSS) in cognitive radio networks (CRNs) enhances spectral decision-making precision but introduces vulnerabilities to malicious secondary user (SU) attacks. This paper proposes a decentralized trust and reputation management (TRM) framework to address these vulnerabilities, emphasizing the need to mitigate risks associated with centralized systems. Inspired by blockchain technology, we present a distributed TRM method for CSS in CRNs, significantly reducing the impact of malicious attacks. Our approach leverages a Proof of Trust (PoT) system to enhance the integrity of CSS, thereby improving the accuracy of spectral decision-making while reducing false positives and false negatives. In this system, SUs’ trust scores are dynamically updated based on their sensing reports, and they will collaboratively participate in new blocks' formation using the trust scores. Simulation results validate the effectiveness of the proposed method, indicating its potential to enhance security and reliability in CRNs.

FedCrow: Federated-Learning-Based Data Privacy Preservation in Crowd Sensing

In the process of completing large-scale and fine-grained sensing tasks for the new generation of crowd-sensing systems, the role of analysis, reasoning, and decision making based on artificial intelligence has become indispensable. Mobile crowd sensing, which is an open system reliant on the broad participation of mobile intelligent terminal devices in data sensing and computation, poses a significant risk of user privacy data leakage. To mitigate the data security threats that arise from malicious users in federated learning and the constraints of end devices in crowd-sensing applications, which are unsuitable for high computational overheads associated with traditional cryptographic security mechanisms, we propose FedCrow, which is a federated-learning-based approach for protecting crowd-sensing data that integrates federated learning with crowd sensing. FedCrow enables the training of artificial intelligence models on multiple user devices without the need to upload user data to a central server, thus mitigating the risk of crowd-sensing user data leakage. To address security vulnerabilities in the model data during the interaction process in federated learning, the system employs encryption methods suitable for crowd-sensing applications to ensure secure data transmission during the training process, thereby establishing a secure federated-learning framework for protecting crowd-sensing data. To combat potential malicious users in federated learning, a legitimate user identification method based on the user contribution level was designed using the gradient similarity principle. By filtering out malicious users, the system reduces the threat of attacks, thereby enhancing the system accuracy and security. Through various attack experiments, the system’s ability to defend against malicious user attacks was validated. The experimental results demonstrate the method’s effectiveness in countering common attacks in federated learning. Additionally, through comparative experiments, suitable encryption methods based on the size of the data in crowd-sensing applications were identified to effectively protect the data security during transmission.

User's Benefit Perceptions and Privacy Concerns for the Personalised Recommendation Service on Taobao: A Privacy Calculus Perspective

First proposed in 1992 by Goldberg, Nichols, Oki, and Terry, Recommendation System (RS) was regarded as an efficient tool for filing electric documents. Personalised recommendations are defined as a personalised list of items that the recommendation algorithm process the massive data to produce results (Wang et al., 2020). After decades of research and development, it has been utilised in fields such as social networks, streaming services, and e-commerce. (Ko et al., 2022). Promotion of customer satisfaction is one of the most mentioned goals of recommendation systems with its capability to deal with overloaded information and find relevant items that may interest customers. According to Abdollahpouri et al. (2020), a recommendation system is economically beneficial and sustainable for service providers and capable of creating values for users and perhaps further stakeholders. To balance customer satisfaction with personalised and targeted services economic-oriented goals like conversion rates, higher revenues, and client retention, personalised recommendations are provided in various contexts (Jannach and Jugovac, 2019). However, a great portion of data processed is collected from users causing risks of privacy leakage and users' concerns about privacy violations (Meng et al., 2018). It is prevailing for users to perceive privacy invasions due to the possibility of data sharing, renting, or selling to third parties (Liu et al., 2017). Malicious user attacks may occur with data collected from users and results from learning these data (Du et al., 2018). Discrimination issues and merchant-beneficial results also arise with the underlying data provided by users (Ekstrand et al., 2019; Ricci et al., 2022). As the Privacy Calculus Theory indicates, interests and privacy risks perceived from the Internet jointly influence users' willingness to personal information disclosure (Dinev and Hart, 2006). Keywords: User's Benefit Perceptions, Privacy Concerns, Privacy Calculus

A Collaborative Spectrum Sensing Algorithm Based on Reputation Update against Malicious User Attacks

In cognitive radio networks, collaborative spectrum sensing (CSS) algorithms could improve spectrum detection performance; however, most explorations are based on reliable network environments. In the real network environment, there may be malicious users that bring wrong spectrum sensing results and attacks designed by them remarkably reduce the spectrum efficiency. In order to resist the attacks of malicious users, this paper proposes a CSS method based on the reputation update. By setting an appropriate reputation threshold, the user fusion center selects the sensing user with a higher reputation to participate in the CSS. Each user’s reputation value is then updated according to whether its local sensing result matches the final judgment result. This article chiefly discusses scenarios for application of three information fusion rules. The simulation results reveal that the proposed approach with reputation update outperforms the conventional CSS algorithm for a variety of judgment rules. The proposed algorithm is capable of preventing lower reputation users from participating in the CSS, filtering out malicious users, and eliminating the impact of malicious users’ attacks.

Secure Threshold Ring Signature Based on SM9

Nowadays, it becomes a major issue that has become increasingly prominent to ensure the privacy and security of information. With the wide application of SM9 algorithm in various fields, it is particularly important to improve the security of SM9 algorithm. Therefore, many researchers use the SM9 algorithm for supporting the underlying cryptography. And group signature, ring signature and two-party signature are introduced in succession to design the related scheme with significant effecting. In order to further improve the signature security and promote the promotion of SM9 algorithm. First, the SM9 identification and cipher algorithm is improved by using multi-KGC(Key Generation Center) to generate system parameters to achieve the purpose of improving the security of the key in this paper. Second, a threshold ring signature scheme based on SM9 is proposed by combining SM9 with threshold ring signature. The analysis shows that, compared with the existing schemes, the proposed scheme has more advantages in security, such as higher anonymity, unforgeability and non-repudiation, while resisting malicious user attacks and malicious KGC attacks. This scheme extends SM9 from single user signature to multi-users signature, and increases the application scenarios of SM9, which plays a positive role in the promotion of SM9.

Three Layer Security Model for Distributed Relational Database

Today we are living in a digital world where we all are connected to an open access network to exchange information. This information is more valuable assets that need protection from unauthenticated access. There are many security models has been proposed to protect the data stored in database but still needs more protection from attacks. In this study we have a three layer security model for relational database. Proposed model builds a security architecture based on access control policy, cryptography techniques and stored procedure. These mechanisms and techniques are combined together to present an interactive three-layer security model for securing relational databases in distributed environment. This model provides a kind of security controlling both in rests of database and data in motion from malicious user and other attacks.

Enhancing the Robustness of Neural Collaborative Filtering Systems Under Malicious Attacks

Recommendation systems have become ubiquitous in online shopping in recent decades due to their power in reducing excessive choices of customers and industries. Recent collaborative filtering methods based on the deep neural network are studied and introduce promising results due to their power in learning hidden representations for users and items. However, it has revealed its vulnerabilities under malicious user attacks. With the knowledge of a collaborative filtering algorithm and its parameters, the performance of this recommendation system can be easily downgraded. Unfortunately, this problem is not addressed well, and the study on defending recommendation systems is insufficient. In this paper, we aim to improve the robustness of recommendation systems based on two concepts— stage-wise hints training and randomness . To protect a target model, we introduce noise layers in the training of a target model to increase its resistance to adversarial perturbations. To reduce the noise layers’ influence on model performance, we introduce intermediate layer outputs as hints from a teacher model to regularize the intermediate layers of a student target model. We consider white box attacks under which attackers have the knowledge of the target model. The generalizability and robustness properties of our method have been analytically inspected in experiments and discussions, and the computational cost is comparable to training a standard neural network-based collaborative filtering model. Through our investigation, the proposed defensive method can reduce the success rate of malicious user attacks and keep the prediction accuracy comparable to standard neural recommendation systems.

Malicious Cognitive User Identification Algorithm in Centralized Spectrum Sensing System

Collaborative spectral sensing can fuse the perceived results of multiple cognitive users, and thus will improve the accuracy of perceived results. However, the multi-source features of the perceived results result in security problems in the system. When there is a high probability of a malicious user attack, the traditional algorithm can correctly identify the malicious users. However, when the probability of attack by malicious users is reduced, it is almost impossible to use the traditional algorithm to correctly distinguish between honest users and malicious users, which greatly reduces the perceived performance. To address the problem above, based on the β function and the feedback iteration mathematical method, this paper proposes a malicious user identification algorithm under multi-channel cooperative conditions (β-MIAMC), which involves comprehensively assessing the cognitive user’s performance on multiple sub-channels to identify the malicious user. Simulation results show under the same attack probability, compared with the traditional algorithm, the β-MIAMC algorithm can more accurately identify the malicious users, reducing the false alarm probability of malicious users by more than 20%. When the attack probability is greater than 7%, the proposed algorithm can identify the malicious users with 100% certainty.

Malicious User Attack in Cognitive Radio Networks

Signal detection in cognitive radio network (CRN) is influenced by several factors. One of them is malicious user that emulate primary user (PU) signal. Emulation of PU signal causes detection error. This paper investigates the impact of malicious user attack to PU signal detection. A number of malicious users are randomly deployed around secondary user (SU) at a certain distance. They attempt to attack primary signal detection that is transmitted from 100 km to SU receiver. Then, the received signal power at secondary receiver and the performance of probability of false alarm and probability of miss detection under two hypothesis of Neyman Pearson criterion are studied. The derived results show that a number of malicious users has a significant impact to the performance of received power at SU and detection error rate.

User Activity Based Application-Layer DoS/DDoS Attack Defense Algorithm

In application-layer DoS/DDoS attacks, malicious users attack the victim server by sending lots of legitimate requesting packages, which overwhelm the server bottleneck resources. Normal user’s request thus may not be satisfied. The traditional intrusion detection systems for network-layer cannot effectively identify this attack, and recent researches on this kind of attack are mainly for Web servers. This paper proposed a new defense algorithm based on user activity for topic-based Pub/Sub communication servers in mobile push notification systems. Users consuming system bottleneck resources the most can get high scores and thus are considered overactive. With some resource retaken strategy, overactive users’ connections will be dropped according to system performance level. Therefore, the system can get rid of latent threatens. Experiments indicated that this algorithm can identify normal and abnormal users well.

Optimisation of cooperative spectrum sensing for cognitive radio networks in the presence of smart primary user emulation attack

AbstractCooperative spectrum sensing, which is an effective method to improve the detection performance in cognitive radio (CR), imposes some threats to the network. One of these common threats is primary user emulation attack, where some malicious users try to mimic the primary signal and deceive secondary users to prevent them from accessing the vacant frequency. Here, we propose a cooperative spectrum sensing scheme for CR network in the presence of smart primary user emulation attack with spectrum sensing capability. More precisely, we optimise main parameters involved in cooperative energy detection spectrum sensing, such as the voting rule, the number of samples used for making the decision and the detection threshold so as to minimise the probability of error when energy detection is corrupted by malicious user's attack. Simulation results indicate performance improvement brought by our proposed method compared with conventional defense techniques in the presence of attacker. Copyright © 2014 John Wiley & Sons, Ltd.

Secure And Reliable Routing In Mobile Adhoc Networks

The growing diffusion of wireless-enabled portable devices and the recent advances in Mobile Ad-hoc NETworks (MANETs) open new scenarios where users can benefit from anywhere and at any time for impromptu collaboration. However, energy constrained nodes, low channel bandwidth, node mobility, high channel error rates, channel variability and packet loss are some of the limitations of MANETs. MANETs presents also security challenges. These networks are prone to malicious users attack, because any device within the frequency range can get access to the MANET. There is a need for security mechanisms aware of these challenges. Thus, this work aims to provide a secure MANET by changing the frequency of data transmission. This security approach was tested, and the results shows an interesting decreased of throughput from malicious node when the number of frequency used is increased, that way the MANET will not waste it’s resources treating malicious packets. The other contribution of this work is a mobility aware routing approach, which aims to provide a more reliable routing by handling effectively the nodes mobility.

Dynamic Identity Based Authentication Protocol for Two-Server Architecture

Most of the password based authentication protocols make use of the single authentication server for user's authentication. User's verifier information stored on the single server is a main point of susceptibility and remains an attractive target for the attacker. On the other hand, multi-server architecture based authentication protocols make it difficult for the attacker to find out any significant authentication information related to the legitimate users. In 2009, Liao and Wang proposed a dynamic identity based remote user authentication protocol for multi-server environment. However, we found that Liao and Wang's protocol is susceptible to malicious server attack and malicious user attack. This paper presents a novel dynamic identity based authentication protocol for multi-server architecture using smart cards that resolves the aforementioned flaws, while keeping the merits of Liao and Wang's protocol. It uses two-server paradigm by imposing different levels of trust upon the two servers and the user's verifier information is distributed between these two servers known as the service provider server and the control server. The proposed protocol is practical and computational efficient because only nonce, one-way hash function and XOR operations are used in its implementation. It provides a secure method to change the user's password without the server's help. In e-commerce, the number of servers providing the services to the user is usually more than one and hence secure authentication protocols for multi-server environment are required.

SECURE ROUTING IN WSN

The growing diffusion of wireless sensors and the recent advances in Wireless Sensor Networks (WSNs) open new scenarios where sensors can be rapidly deployed without any existing infrastructure. Such networks are useful in many fields, such as emergency rescue, disaster relief, smart homes systems, patient monitoring, industrial applications, health monitoring, environmental control, military applications, etc. However WSN presents many challenges. These networks are prone to malicious users attack, because any device within the frequency range can get access to the WSN. There is a need for security mechanisms aware of the sensor challenges (low energy, computational resources, memory, etc.). Thus, this work aims to provide a secure WSN by changing the frequency of data transmission. This security approach was tested, and the results shows an interesting decreased of throughput from malicious node when the number of frequency used is increased, that way the WSN will not waste it’s resources treating malicious packets.