Abstract Cardiovascular disease (CVD) believes to be a major cause of transience and indisposition worldwide. Early diagnosis and timely intervention are critical in preventing the progression of CVD and improving patient outcomes. Machine learning (ML) algorithms have emerged as powerful tools in CVD recognition, with the potential to assist physicians in making accurate and efficient diagnoses. This research paper explores the combination of multiple ML algorithms for CVD recognition, utilizing diverse datasets such as the Cleveland, Hungarian, Switzerland, statlog, and VA Long Beach datasets. Additionally, a CVD dataset comprising 12 attributes and 70,000 records is employed, demonstrating improved results through the proposed and trained model compared to previous prediction techniques for CVD. The performance of various ML techniques, including support vector machines (SVM), naive Bayes (NB), K‐nearest neighbour (KNN), random forest (RF), and logistic regression (LR), is evaluated and compared. The impact of feature selection and feature scaling on the models' performance is also examined. An ensemble bagging technique is applied which is being embedded with other classifiers. LR classifier embedded with bagging techniques proved to be our proposed model. The findings reveal that the proposed Hybrid Linear Regression Bagging Model (HLRBM) outperforms other models. Furthermore, the study highlights the significance of data preprocessing techniques, such as data normalization and class balancing, which significantly enhance the performance of all models. To this end, standard scalar and synthetic minority over‐sampling technique (SMOTE) are employed. The study emphasizes the importance of selecting an appropriate ensemble technique in conjunction with various ML algorithms and preprocessing methods for CVD prediction. Overall, the research provides valuable insights into the potential of ML in improving CVD risk assessment.

Abstract Fog's basic distributed nature and ability to process data in transit—that is, to make decisions in real time—make it a good fit for scenarios involving several distributed devices that need to communicate, provide real‐time data analysis, and carry out storage functions. The majority of fog computing applications are driven by the user's demands and/or their desire for functioning services, either neglecting or giving security considerations second attention. Fog computing security issues have not received enough attention. Fog computing could be exploitable due to the security difficulties associated with cloud computing. Due to its flexibility to function near the end user and independence from a centralized design, fog computing provides the dependability required by time‐sensitive smart healthcare systems. There is a need for enhanced security and privacy solutions for fog computing, where trust is essential, due to the importance of healthcare data. This research aims to develop a context‐based adaptive trust solution for the smart healthcare environment utilizing Bayesian approaches and similarity measures against bad mouthing and ballot stuffing, while context‐dependent trust solutions for fogs remain an unexplored area of study. The proposed trust model has been simulated in Contiki‐Cooja to evaluate our findings. In contrast to static weighting, adaptive weights are provided to direct and indirect trust using entropy values that ensure the least degree of trust bias, and context similarity calculations eliminate recommender nodes with malicious intent by leveraging server, colleague, and service similarities. The proposed model protects smart healthcare systems from attacks using similarity metrics, incorporates context, and also uses adaptive weighting for trust calculation. By eliminating trust bias and also detecting attacks, this solution enhances the trust calculation by 10% as compared to the previous solution. This paradigm is efficient due to its small trust computation overhead and linear complexity O ( n ).

Social media platforms are heavily used by people to express their views in their native languages. Besides positive views, people often use abusive or offensive language to express their anger or frustration. Resource-rich languages have offensive language detection systems which automatically monitor and block offensive content, however, they are very rare for low-resourced languages. This is because of the nonavailability of datasets for local languages. This article proposes a model which automatically detects offensive language for a very low-resource language, i.e., Pashto. The Roman Pashto dataset is created by picking 60 thousand comments from different social media and labeling them manually. The proposed model is trained and tested using three different feature extraction approaches, i.e., bag-of-words (BoW), term frequency-inverse document frequency (TF-IDF), and sequence integer encoding. Four traditional classifiers and a deep sequence model are used to train on this task. Experimental result shows that the random forest classifier works best and give 94.07 % testing accuracy on a combination of unigrams, bigrams, and trigrams. The same classifier gives maximum accuracy of 93.90 % with TF-IDF. However, the overall highest testing accuracy of 97.21% is achieved by using bidirectional long short-term memory (BLSTM). The corpus created in this work is made available for the researcher working in this domain.

Fog’s inherent decentralized nature and ability to process data in transit, i.e., the ability to draw conclusions in real-time, are quite suitable for scenarios where an enormous number of decentralized devices need to communicate and provide live analysis of data and storage tasks. Fog computing’s ability to work close to the end user and non-reliance on centralized architecture provides the dependability that time-critical smart healthcare systems need. Because of the critical nature of healthcare data, better security and privacy solutions for fog computing are required, with trust being of the utmost importance. Context-dependent trust solution for fogs is still an open research area, so the aim of this research is to propose a context-based adaptive trust solution for smart healthcare environments using a Bayesian approach and similarity measures. The proposed trust model has been simulated in Contiki, Cooja, and a Java-based application has been developed to analyze our results. Adaptive weights assigned to direct and indirect trust using entropy values ensure the minimization of trust bias as opposed to static weighting. Context-based similarity calculations filter out recommender nodes with malicious intent using server, social contact, and service similarity. This model is efficient and has a low-trust computation overhead because it has a linear complexity of$O(n)$.

Physical objects are getting connected to the Internet at an exceptional rate, making the idea of the Internet of Things (IoT) a reality. The IoT ecosystem is evident everywhere in the form of smart homes, health care systems, wearables, connected vehicles, and industries. This has given rise to risks associated with the privacy and security of systems. Security issues and cyber attacks on IoT devices may potentially hinder the growth of IoT products due to deficiencies in the architecture. To counter these issues, we need to implement privacy and security right from the building blocks of IoT. The IoT architecture has evolved over the years, improving the stack of architecture with new solutions such as scalability, management, interoperability, and extensibility. This emphasizes the need to standardize and organize the IoT reference architecture in federation with privacy and security concerns. In this study, we examine and analyze 12 existing IoT reference architectures to identify their shortcomings on the basis of the requirements addressed in the standards. We propose an architecture, the privacy-federated IoT security reference architecture (PF-IoT-SRA), which interprets all the involved privacy metrics and counters major threats and attacks in the IoT communication environment. It is a step toward the standardization of the domain architecture. We effectively validate our proposed reference architecture using the architecture trade-off analysis method (ATAM), an industry-recognized scenario-based approach.

Purpose This article aims to provide an interactive model for hands on training of malware analysis. Cyberwar games are the highly stylized representation of cyber conflicts in a simulation model. Game models are helpful in understanding the phenomenon of cyber attacks as well as to evolve new techniques of detection. Cyber security trainings are generally very challenging. Cyber test beds make such trainings easy both for trainees and trainers. However, it is not feasible for each organization to build a network for the sole purpose of hands-on training for employees. Therefore, it is desirable to build an interactive environment that is interesting and free of cost as well. Design/methodology/approach After exploring existing cyberwar games and their techniques, limitation and strengths, this paper presents a design to merge the cyber attacks into a unique model of war game for detection and analysis of malware. This research designs a malware analysis testbed using online free resources. The authors have used the platform of Cyber Defense Technology Experimental Research (DETER). This study proposed model of a testbed that supports malware reverse engineering scenarios, exercise logs and analysis to develop reverse engineering tactics. Findings The proposed cyber testbed is an automated system that can be used as a platform to train cyber warriors. A few features of the proposed testbed are as follows: testbed provides real or seemingly real malware communication with the real world. It includes automated decisions for the detection of malicious behavior without human intervention. The author gives a design to develop free of cost mechanism for remote learning of highly technical cyber security areas, and this simulation is for malware analysis. Originality/value Cyberwar games are built for strengthening offensive and defensive capabilities in cyberspace. For this purpose, many simulations, as well as emulation platforms, can be found. Some are free and open-source, whereas others are commercial and quite expensive. Existing testbeds have limitations in respect of cyberwar games for creating innovative scenarios. Existing literature does not offer any attack and response scenario developed for malware detection through some existing open-source and online simulation or emulation environments. This research includes an analysis of the existing platforms as well as the design of a new model for malware analysis and training.

The emergence of software-defined networking (SDN) has revolutionized the management of an enterprise network. The SDN-based design provides flexibility in network management, which spans over multiple applications, e.g., routing, switching, forwarding, and controlling. It reduces the reliance on vendor-specific devices and middlebox solutions, such as firewalls, intrusion detection systems (IDSs), intrusion prevention systems (IPSs), etc. Furthermore, due to the integration of different technologies, privacy is one of the core issues faced by the enterprise. Host anonymity is one of the techniques to safeguard against privacy attacks; however, the existing anonymization solutions provide better anonymity, but at the cost of higher latency and are most suited for Internet traffic. To tackle this issue in an enterprise network, we propose an SDN-based communication framework using enterprise integration patterns (EIPs) that offers anonymous communication in an enterprise environment. Host anonymity is achieved by replacing the real IP address with the spoofed IP address during the transmission of data packets inside the network. Unlike the traditional networks, SDN can modify the header fields of packets as they traverse in the network from the source to the destination. In addition to the host anonymity, this framework also provides context-aware communication by leveraging the SDN global visibility characteristic, where application services are discoverable on the network without disclosing the addresses of the application servers. Moreover, context-aware services enable network traffic to be routed based on the application-layer services rather than the network-layer information. In the end, evaluation of the proposed framework is carried out with respect to the performance of anonymous communication, computational complexity, and security of the complete proposed framework. In addition, we also highlighted that the proposed framework is more suitable for heterogeneous network environments such as Internet of Things-based solutions.

The smart connected devices are the first choice of cybercriminals for spreading spy wares and different security attacks. The current security standards and protocols for Internet of Things (IoT) have failed in providing security to these devices. In addition, IoT market giants are producing nonsecure smart products in order to grab the open market. Furthermore, low resources of IoT devices, limits the traditional host-based protection solutions like anti-virus, IDS, IPS, etc. To overcome the resource constraintness and security barriers of smart devices, a network-level security architecture based on lightweight cryptographic parameters is required. Software-defined networking (SDN) is a new networking paradigm to overcome the control, management, and security issues in traditional networking. The SDN controller handles all the computation and complexities at the network level, rather than smart devices. In this research, we first present a new privacy-preserving security architecture for SDN-based smart homes. Subsequently, an anonymous lightweight authentication mechanism (ALAM) is designed based on the proposed security architecture core foundations. Furthermore, the security characteristics of the proposed protocol are formally analyzed using Burrows-Abadi-Needham (BAN) logic and ProVerif, followed by informal security analysis. Finally, performance evaluation and comparative analysis of the scheme is carried out.

In today's era, large data centers are drawn toward the two popular technologies, i.e., enterprise integration patterns (EIP) and software defined networking (SDN). The former is the combination of design patterns which integrates the new and existing business applications in an enterprise environment, whereas the latter is a rapidly evolving networking paradigm that has reshaped the large enterprise network management by introducing the programmable planes and centralized control. The promising features of EIP, i.e., asynchronous communication, reliability and that of SDN, namely, robustness, network programmability, agility, and global visibility can be merged in order to cope with growing network demands and security. In this article, we propose a communication framework that incorporates both EIP and SDN in an enterprise environment. The architecture of the propose communication framework consists of adaptive virtual local area network (VLAN) module to install and delete VLANs reactively using EIP. Furthermore, to enable communication between applications from different networks in an enterprise environment, this framework also contains a packet forwarding module where hosts IP addresses are concealed from each other. In the end, we demonstrate through simulations that how the proposed integrated design can be helpful in improving the security, efficiency, and reliability aspects of the enterprise network. As proof of concept, we also discuss the deployment of our proposed framework in IoT-based Smart Cities, which is the extension of EIP on a large scale.

Smartphone users spend a substantial amount of time in browsing, emailing, and messaging through different social networking apps. The use of social networking apps on smartphones has become a dominating part of daily lives. This momentous usage has also resulted in a huge spike in cybercrimes such as social harassing, abusive messages, vicious threats, broadcasting of suicidal actions, and live coverage of violent attacks. Many of such crimes are carried out through social networking apps; therefore, the forensic analysis of allegedly involved digital devices in crime scenes and social apps installed on them can be helpful in resolving criminal investigations. This research is aimed at performing forensic investigation of five social networking apps, i.e., Instagram, LINE, Whisper, WeChat, and Wickr on Android smart phones. The essential motivation behind the examination and tests is to find whether the data resides within the internal storage of the device or not after using these social networking apps. Data extraction and analysis are carried out using three tools, i.e., Magnet AXIOM, XRY, and Autopsy. From the results of these experiments, a considerable amount of essential data was successfully extracted from the examined smartphone. This useful data can easily be recovered by forensic analysts for future examination of any crime situation. Finally, we analyzed the tools on the basis of their ability to extract digital evidences from the device and their performance are examined with respect to NIST standards.

In federation of cloud, multiple cloud service providers share their resources based on certain assumptions and trust. Most of the times, identity of a user and permission to give access are shared using service access requirements. In intercloud security, building mutual trust relationship between multiple cloud federated identities is the main goal. Securing the components involved in cloud federation can be helpful in building mutual trust relationship between federated identities. Intrusion Detection and Prevention techniques helped a lot in detecting the malicious activities performed by the intruders. In this domain of securing data, a lot of research is being done. Recently, artificial intelligence and machine learning have greatly attracted the attention of researchers to integrate the concepts of network security with artificial intelligence. In this research, we have studied artificial intelligence techniques and finalized artificial neural network (ANN) model to detect the intrusions based on anomalies. In previous studies, it was discussed that major challenges in anomaly based intrusion detection systems is to lessen the false positive rate (FPR). So, in this research the main emphasis is done on reducing the failure rate of FPR of the intrusions done to the system while maintaining the overall perfromance and accuracy of the proposed model.

Internet of Things (IoT) is transforming everyone's life by providing features, such as controlling and monitoring of the connected smart objects. IoT applications range over a broad spectrum of services including smart cities, homes, cars, manufacturing, e-healthcare, smart control system, transportation, wearables, farming, and much more. The adoption of these devices is growing exponentially, that has resulted in generation of a substantial amount of data for processing and analyzing. Thus, besides bringing ease to the human lives, these devices are susceptible to different threats and security challenges, which do not only worry the users for adopting it in sensitive environments, such as e-health, smart home, etc., but also pose hazards for the advancement of IoT in coming days. This article thoroughly reviews the threats, security requirements, challenges, and the attack vectors pertinent to IoT networks. Based on the gap analysis, a novel paradigm that combines a network-based deployment of IoT architecture through software-defined networking (SDN) is proposed. This article presents an overview of the SDN along with a thorough discussion on SDN-based IoT deployment models, i.e., centralized and decentralized. We further elaborated SDN-based IoT security solutions to present a comprehensive overview of the software-defined security (SDSec) technology. Furthermore, based on the literature, core issues are highlighted that are the main hurdles in unifying all IoT stakeholders on one platform and few findings that emphases on a network-based security solution for IoT paradigm. Finally, some future research directions of SDN-based IoT security technologies are discussed.

Healthcare industry is continuously evolving as new technologies are being integrated in the existing healthcare infrastructure. Privacy preservation provides one of the key role from security perspective in an electronic healthcare data security. Due to continuously emerging threats, a comprehensive security framework that should provide strong patient anonymity level (considering both patient's identity and patient's data), anonymized data searching and successful correlation of PHR for medical research is difficult to formulize. In this paper, some previous security frameworks were analyzed and their flaws were identified. Also it proposed an enhancement to the existing solutions based on a careful analysis of their shortcomings.

Cloud data sharing is a prevalent topic these days due to its immense benefits. But regardless of this, industry is facing a huge setback due to deficiencies in present security frameworks. Achieving security becomes favorable if we put together a strong authentication mechanism. Authentication techniques as much random and complex they get are still breakable, so a truly random and strong technique is required. For this purpose, a new decentralized approach consisting of multiple authentication factors (username and password, biometrics and tokens) is being proposed for authentication which gives strong foundation of security and is difficult to break. The analysis showed significantly better comparative results.

Cloud computing is the emerging field nowadays and it has truly revolutionized the domain of Information Technology. This domain is very large and not easy to handle especially when it comes to the forensic in a cloud environment that is considered a very cumbersome process. This paper presents a feasibility analysis of performing digital forensics via SIEM (Security Information and Event Management) system in cloud environment. The research work mainly focuses on passive attacks while some active attacks are also covered and the forensics analysis is done while considering the service provider end. The preliminary analysis presented in this paper will provide a comprehensive overview of the various artifacts that may be considered for performing an in-depth forensic analysis in cloud environment using Security Information and Event Management System.

Cloud computing has appeared to be state of the art in the world of Information Technology especially in healthcare systems due to its innovative computing deployment model as a source of utility for users. However, many healthcare organizations are disinclined to adapt cloud computing due to security shortcoming associated with its infrastructure and many of them are still unaddressed. To provide a promising security to cloud computing infrastructure is a key issue, as users have a tendency to store their sensitive and classified medical data with cloud service providers which include the both trusted and un-trusted parties of cloud service providers. This paper addresses the distinguishing features of cloud such as, rapid elasticity, pooling of resources, on-demand services to users, multi-tenancy, third-party rule and its security requirements. Then, it presents the analysis of security concerns of cloud computing for healthcare systems in terms of availability, trust, confidentiality, compliance, integrity and audit. Furthermore, single-cloud is far more vulnerable to failure of service unavailability and malicious insiders and due to this reason it is less popular in healthcare, as medical healthcare systems are concerned about its security. From this notion of security concern an advanced model has emerged; “multi-cloud” also known to be “cloud-of-clouds”. This paper also provides an insight of security in single-cloud and multi-cloud and possible security recommendations for healthcare systems, as it has been observed that single-cloud has received more focus from researchers in terms of security vulnerabilities than a multi-cloud environment.