Federated learning (FL) is a privacy-preserving machine learning technique that facilitates collaboration among participants across demographics. FL enables model sharing, while restricting the movement of data. Since FL provides participants with independence over their training data, it becomes susceptible to poisoning attacks. Such collaboration also propagates bias among the participants, even unintentionally, due to different data distribution or historical bias present in the data. This paper proposes an intentional fairness attack, where a client maliciously sends a biased model, by increasing the fairness loss while training, even considering homogeneous data distribution. The fairness loss is calculated by solving an optimization problem for fairness metrics such as demographic parity and equalized odds. The attack is insidious and hard to detect, as it maintains global accuracy even after increasing the bias. We evaluate our attack against the state-of-the-art Byzantine-robust and fairness-aware aggregation schemes over different datasets, in various settings. The empirical results demonstrate the attack efficacy by increasing the bias up to 90%, even in the presence of a single malicious client in the FL system.

Due to its admirable accuracy and performance across a wide range of classification and identification tasks, deep learning algorithms have gained popularity in several applications. However, the models’ security has become a serious concern, as antagonists could use them to promote their malicious goals. This work proposes a triggerless targeted model poisoning attack (TriMPA) against deep neural network without requiring any change in input to trigger the backdoor. TriMPA identifies active neurons that highly contribute to the prediction of the victim output label and replaces those neurons with that corresponding to the target output label. The performance of the proposed mechanism is evaluated through experiments as well as analyzed theoretically. It is shown that TriMPA achieves a higher attack success rate.

Federated learning is a collaborative learning paradigm that brings the model to the edge for training over the participants' local data under the orchestration of a trusted server. Though this paradigm protects data privacy, the aggregator has no control over the local data or model at the edge. So, malicious participants could perturb their locally held data or model to post an insidious update, degrading global model accuracy. Recent Byzantine-robust aggregation rules could defend against data poisoning attacks. Also, model poisoning attacks have become more ingenious and adaptive to the existing defenses. But these attacks are crafted against specific aggregation rules. This work presents a generic model poisoning attack framework named Sine (Similarity is not enough), which harnesses vulnerabilities in cosine similarity to increase the impact of poisoning attacks by 20-30%. Sine makes convergence unachievable by maintaining the persistence of the attack. Further, we propose an effective defense technique called FLTC (FL Trusted Coordinates) to avoid such issues. FLTC selects the trusted coordinates and aggregates them based on the change in their direction and magnitude with respect to a trusted base model update. FLTC could successfully defend against poisoning attacks, including adaptive model poisoning attacks, by restricting the attack impact to 2-4%.

The payment channel network (PCN) is an effective off-chain scaling solution widely recognized for reducing operational costs on permissionless blockchains. However, it still faces challenges such as lack of flexibility, channel exhaustion, and poor sustainability. Currently, a separate deposit is required for each payment channel, which locks a significant amount of coins for a longer period. This restricts the ability to move locked coins across their channels off-chain. Additionally, unbalanced (unidirectional) transfers can lead to channel exhaustion, rendering the channel unsustainable. To address these issues, we propose a novel off-chain flexible PCN protocol called Flexichain . Unlike existing approaches, Flexichain allows users to deposit coins per user rather than per channel. This provides flexibility to move coins freely between channels without relying on the blockchain or disrupting the off-chain cycle. Flexichain is proven to be secure under the Universal Composability framework and resistant against channel exhaustion attacks. To assess the performance of Flexichain, we conduct experiments on both on-chain and off-chain operations using snapshots of the Lightning Network. We evaluate the on-chain gas costs, success ratio and success amount of off-chain payments under uniform and skewed payment demands, as well as the computational and communication overheads of the off-chain contracts.

In breast cancer treatment, accurately predicting how long a patient will survive is crucial for decision-making. This information guides treatment choices and supports patients' psychological recovery. To address this challenge, we introduce a novel predictive model to forecast breast cancer prognosis by leveraging diverse data sources, including clinical records, copy number variation, gene expressions, DNA methylation, microRNA (miRSeq) sequencing, and whole slide image data from the TCGA Database. The methodology incorporates graph contrastive learning with cross-modality attention (CAGCL), considering all possible combinations of the six distinct data modalities. Feature embeddings are enhanced through graph contrastive learning, which identifies subtle differences and similarities among samples. Further, to learn the complementary nature of information across multiple data modalities, a cross-attention framework is proposed and applied to the graph contrastive learning-based extracted features from various data sources for breast cancer survival prediction. It performs a binary classification to anticipate the likelihood of short- and long-term breast cancer survivors, delineated by a five-year threshold. The proposed model (CAGCL) showcases superior performance compared to baseline models and other state-of-the-art models. The model attains an accuracy of 0.932, a sensitivity of 0.954, a precision of 0.958, an F1 score of 0.956, and an AUC of 0.948, underscoring its effectiveness in predicting breast cancer survival.

Machine learning (ML) has seen widespread adoption across different domains and is used to make critical decisions. However, with profuse and diverse data available, collaboration is indispensable for ML. The traditional centralized ML for collaboration is susceptible to data theft and inference attacks. Federated learning (FL) promises secure collaborative machine learning by moving the model to the data. However, FL faces the challenge of data and model poisoning attacks. This is because FL provides autonomy to the participants. Many Byzantine-robust aggregation schemes exist to identify such poisoned model updates from participants. But, these schemes require raw access to the local model updates, which exposes them to inference attacks. Thus, the existing FL is still insecure to be adopted.This paper proposes the very first generic FL framework, which is both resistant to inference attacks and robust to poisoning attacks. The proposed framework uses hyperdimensional computing (HDC) coupled with FL, called HDFL. HDFL is compatible with different (ML) model architectures and existing Byzantine-robust defenses. HDFL restricts drop in accuracy to 1-2%. HDFL does not add any additional communication overheads and incurs negligible computational time in encoding and decoding raw local model updates. Empirical evaluation demonstrates the effectiveness of HDFL. HDFL performs secure aggregation and achieves no-attack accuracy, even in the presence of 40% attackers, in just 1.2s per iteration.

With the emergence of the Internet of Things (IoT), there come opportunities to connect people, data, and objects, bringing dynamic changes in our way of living. At the same time, the presence of sensors and other devices around us has created a leaky ecology that is webbed. It has turned into a risk for the users and causes worry about how widely it could be adopted. Numerous projects have been put up in a similar vein, creating innovations to improve security. But, conventional cryptographic solutions are difficult to embed as these IoT devices have limited resources. Physical unclonable functions (PUFs), particularly in resource-restraining devices, have shown to be valuable for developing authentication protocols. In this work, we propose a PUF-based authentication mechanism for IoT devices. This authentication protocol also performs the key exchange between the two nodes, with the server in between. The security features of the proposed mechanism are verified using AVISPA tool.

Because of the widespread presence and ease of access to the Internet, edge devices are the perfect candidates for providing quality training on a variety of applications. However, their participation is restrained due to potential leakage of sensitive and private data. Federated learning targets to address these issues by bringing the model to the device and keeping the data in place. Still, it suffers from inherent security issues such as malicious participation and unfair contribution. The central server may become a bottleneck as well as induce biased aggregation and incentives. This article proposes a blockchain assisted federated learning framework, which fosters honest participation with reduced overheads, facilitating fair contribution-based weighted incentivization. A new consensus mechanism named PoIS (Proof of Interpretation and Selection) is proposed based on honest clients’ contributions. PoIS uses model interpretation techniques for evaluating and calculating individual contributions. The aggregation of feature attributions in PoIS, is able to detect the adversaries, and the label-wise aggregation of attributions across the participants helps to define the prominent contributors. Further, we devise a credit function based on the contribution, relevance as well as the past performance for calculating incentives. Extensive experiments have been carried out for the proposed architecture with different settings, models, and datasets, to verify our claim. It successfully restricts the attack to less than 5%, and selects the prominent (top-${k}$) contributors. Theoretical analysis provides the guarantee for byzantine-robust aggregation, in a malicious setting.

Industrial Internet of things (IIoT) refers to a network of smart devices, equipped with a variety of sensors connected to the Internet. The devices in IIoT can be shared among multiple public and/or private applications. These applications can simultaneously access the data generated by these devices, necessitating concurrent data-collection. With devices being power-constrained, the chances of device failures are high in shared device infrastructure. This results in partitioned network topology and impacts data collection. Furthermore, the network-topology reconstruction process is also energy-consuming. This paper proposes a fault-tolerant design of concurrent data col-lection process in shared IIoT applications. Via simulation, we show our proposed algorithm handles device failures without affecting the overall time-duration of concurrent data-collection and handles the faults better as compared to an existing algorithm in terms of better overall data collection time.

Social Internet of Vehicle (SIoV) has emerged as one of the most promising applications for vehicle communication, which provides safe and comfortable driving experience. It reduces traffic jams and accidents, thereby saving public resources. However, the wrongly communicated messages would cause serious issues, including life threats. So it is essential to ensure the reliability of the message before acting on considering that. Existing works use cryptographic primitives like threshold authentication and ring signatures, which incurs huge computation and communication overheads, and the ring signature size grew linearly with the threshold value. Our objective is to keep the signature size constant regardless of the threshold value. This work proposes MuSigRDT, a multisignature contract based data transmission protocol using Schnorr digital signature. MuSigRDT provides incentives, to encourage the vehicles to share correct information in real-time and participate honestly in SIoV. MuSigRDT is shown to be secure under Universal Composability (UC) framework. The MuSigRDT contract is deployed on Ethereum's Rinkeby testnet.

The prediction of the protease cleavage site of an amino acid sequence of Human Immune Deficiency Virus (HIV-1) type 1 has various significant applications in finding novel drug targets, Protease Inhibitor study, and many more. HIV-1 protease cleavage site prediction studies have been carried out for decades using different feature extraction techniques by many researchers. Existing studies only focus on the simple representation of features, the latent feature space which is capable of reconstructing the data well was never explored. Motivated by this, we have utilized the concept of Stacked Autoencoder (SAE) for latent feature extraction from the con-catenated feature set which is constructed with three different types of extracted features namely structural, physicochemical and sequential features. The hidden nodes in deep layers of SAE manage the high level abstraction for dimensionality reduction and maintain the key information of the data simultaneously. Several classifiers are trained using this informative feature set and finally majority voting based classifier ensemble technique is applied for cleavage site prediction task. The proposed approach is applied on several benchmark data sets and average accuracy, precision, recall, and F-measure values of 0.95, 0.91, 0.83, 0.86, respectively, are attained. Comparison with respect to other existing studies also establishes the efficacy of the proposed technique.

Social Internet of Vehicles (SIoV) is an emerging technology in the smart city environment, enabling smart vehicles to form social groups and exchange data among themselves. SIoV facilitates many applications aiming to improve driving safety and traffic monitoring by sharing data among vehicles. It ensures a safe and comfortable drive. However, privacy, data confidentiality, and data integrity are the major challenges during multi-hop data transfer that must be addressed for the wide adoption of SIoV. The existing solutions do not provide anonymity and consume more network resources. To address these issues, we propose SIoVChain, a time-lock contract-based privacy-preserving data sharing scheme with incentives for SIoV. It does not only enable data sharing between vehicles anonymously but incentivizes them also. In addition, the proposed framework imposes a penalty anonymously if a malicious vehicle disseminates false information. SIoVChain is shown to be secure against stealing processing fee attack while preserving sender-receiver privacy and path privacy. Universal Composability (UC) framework is used to verify user privacy. The feasibility and efficiency of the scheme are also demonstrated.

The tremendous popularity of online social network (OSN) services in recent years has offered a new way of information sharing. OSN services allow the spread of information like fire in a forest, among the target audience. On the other hand, these services raise serious concerns about the privacy of their users. In this article, we propose a novel exponential model to measure the visibility of tweets exploiting the trust and interest of the followers. Furthermore, we develop a polynomial regression model and a deep neural network (DNN) model for visibility prediction. The experimental results with the real data show that the exponential model achieves better visibility prediction accuracy in comparison to the regression model and DNN model for a specific value of hop count.

Clinical trials and drug discovery would not be effective without the collaboration of institutions. Earlier, it has been at the cost of individual’s privacy. Several pacts and compliances have been enforced to avoid data breaches. The existing schemes collect the participant’s data to a central repository for learning predictions as the collaboration is indispensable for research advances. The current COVID pandemic has put a question mark on our existing setup where the existing data repository has proved to be obsolete. There is a need for contemporary data collection, processing, and learning. The smartphones and devices held by the last person of the society have also made them a potential contributor. It demands to design a distributed and decentralized Collaborative Learning system that would make the knowledge inference from every data point. Federated Learning [21], proposed by Google, brings the concept of in-place model training by keeping the data intact to the device. Though it is privacy-preserving in nature, however, it is susceptible to inference, poisoning, and Sybil attacks. Blockchain is a decentralized programming paradigm that provides a broader control of the system, making it attack resistant. It poses challenges of high computing power, storage, and latency. These emerging technologies can contribute to the desired learning system and motivate them to address their security and efficiency issues. This article systematizes the security issues in Federated Learning, its corresponding mitigation strategies, and Blockchain’s challenges. Further, a Blockchain-based Federated Learning architecture with two layers of participation is presented, which improves the global model accuracy and guarantees participant’s privacy. It leverages the channel mechanism of Blockchain for parallel model training and distribution. It facilitates establishing decentralized trust between the participants and the gateways using the Blockchain, which helps to have only honest participants.

Cloud computing is an emergent paradigm to provide reliable and resilient infrastructure enabling the users (data owners) to store their data and the data consumers (users) can access the data from cloud servers. This paradigm reduces storage and maintenance cost of the data owner. At the same time, the data owner loses the physical control and possession of data which leads to many security risks. Therefore, auditing service to check data integrity in the cloud is essential. This issue has become a challenge as the possession of data needs to be verified while maintaining the privacy. To address these issues this work proposes a secure and efficient privacy preserving provable data possession (SEPDP). Further, we extend SEPDP to support multiple owners, data dynamics and batch verification. The most attractive feature of this scheme is that the auditor can verify the possession of data with low computational overhead.

Fog computing architecture allows the end-user devices of an Internet of Things (IoT) application to meet their latency and computation requirements by offloading tasks to a fog node in proximity. This fog node in turn may offload the task to a neighboring fog node or the cloud-based on an optimal node selection policy. Several such node selection policies have been proposed that facilitate the selection of an optimal node, minimizing delay and energy consumption. However, one crucial assumption of these schemes is that all the networked fog nodes are authorized part of the fog network. This assumption is not valid, especially in a cooperative fog computing environment like a smart city, where fog nodes of multiple applications cooperate to meet their latency and computation requirements. In this paper, we propose a secure task-offloading framework for a distributed fog computing environment based on smart-contracts on the blockchain. The proposed framework allows a fog-node to securely offload tasks to a neighboring fog node, even if no prior trust-relation exists. The security analysis of the proposed framework shows how non-authenticated fog nodes are prevented from taking up offloading tasks.

Intrusion Detection Systems (IDS) has become an indespensive tool to protect the network by detecting the attacks. As each attack is associated with a large number of attributes, it is challenging to select a good set of attributes for achieving better classification accuracy. This work proposes an effective unsupervised feature selection technique using Genetic algorithm (GA) for analyzing IDS data. The search capability of GA has been utilized for optimizing different unsupervised feature quality measures including Pearson correlation, mutual information, and entropy. Different combinations of these features are utilized as fitness functions of the proposed GA based framework. The algorithm is able to find out that subset of features which are uncorrelated and mutually exclusive to each other. Finally, the optimal feature subset obtained is utilized for developing classification systems using some popular machine learning models like decision trees, support vector machines, k-nearest neighbor classifier on the KDD-Cup 99 dataset. The experimental results show that decision tree produces better results than other classifiers. The result confirms 99.62% accuracy, 98.78% detection rate and 0.25% false alarm rate. The most attractive feature of the proposed scheme is that it does not require any labeled information during the feature selection process.

Pastry is a popular DHT overlay for its capability to facilitate storage and retrieval of data/ file among large number of nodes, efficiently without any centralized server. Performance is one of the major concerns in such networks. This work, frames a collaborative attack against Pastry named CAP to reduce the performance. Malicious nodes in CAP, intelligently poison the routing tables by sending fake state-table update messages to a target node. The malicious nodes drop (or forwards to another malicious node) the look-up requests received from a peer. Impact of CAP is observed and found to be significant degradation in performance. Further, we propose a defense mechanism to reduce the effects of such attack.

To achieve security in wireless sensor networks (WSN), communications between sensor nodes need to be encrypted and authenticated. Therefore, keys for encryption and authentication must be agreed among the communicating nodes. At the same time, small memory, weak processor and limited battery power of a sensor node are the major obstacles to implement the traditional security primitives. Owing to both the requirement and obstacles, this paper proposes an efficient key establishment protocol that establishes keys between $\tau$ number of its communicating nodes at a single instance, and therefore the proposed scheme reduces the execution cost.