Diabetic retinopathy (DR) is a worldwide problem associated with the human retina. It leads to minor and major blindness and is more prevalent among adults. Automated screening saves time of medical care specialists. In this work, we have used different deep learning (DL) based 3D convolutional neural network (3D-CNN) architectures for binary and multiclass (5 classes) classification of DR. We have considered mild, moderate, no, proliferate, and severe DR categories. We have deployed two artificial data augmentation/enhancement methods: random weak Gaussian blurring and random shift along with their combination to accomplish these tasks in the spatial domain. In the binary classification case, we have found the performance of 3D-CNN architecture trained by deploying combined augmentation methods to be the best, while in the multiclass case, the performance of model trained without augmentation is the best. It is observed that the DL algorithms working with large volumes of data may achieve better performances as compared to the methods working with small volumes of data.

The Industrial Internet of Things (I-IoT) is a manifestation of an extensive industrial network that interconnects various sensors and wireless devices to integrate cyber and physical systems. While I-IoT provides a considerable advantage to large-scale industrial enterprises, it is prone to significant security challenges in the form of sophisticated attacks such as Advanced Persistent Threat (APT). APT is a serious security challenge to all kinds of networks, including I-IoT. It is a stealthy threat actor, characteristically a nation-state or state-sponsored group that launches a cyber attack intending to gain unauthorized access to a computer network and remain undetected for a longer period. The latest intrusion detection systems face several challenges in detecting such complex cyber attacks in multifarious networks of I-IoT, where unpredictable and unexpected cyber attacks of such sophistication can lead to catastrophic effects. Therefore, these attacks need to be accurately and promptly detected in I-IoT. This paper presents an intelligent APT detection and classification system to secure I-IoT. After pre-processing, several machine learning algorithms are applied to detect and classify complex APT signatures accurately. The algorithms include Decision Tree, Random Forest, Support Vector Machine, Logistic Regression, Gaussian Naive Bayes, Bagging, Extreme Gradient Boosting and Adaboost, which are applied on a publicly available dataset KDDCup99. Moreover, a comparative analysis is conducted among ML algorithms to select the appropriate one for the targeted domain. The experimental results indicate that the Adaboost classifier outperforms the others with 99.9% accuracy with 0.012 s execution time for detecting APT attacks. Furthermore, results are compared with state-of-the-art techniques that depict the superiority of the proposed system. This system can be deployed in mission-critical scenarios in the I-IoT domain.

No abstract

Security has always been the main concern for the internet of things (IoT)-based systems. Blockchain, with its decentralized and distributed design, prevents the risks of the existing centralized methodologies. Conventional security and privacy architectures are inapplicable in the spectrum of IoT due to its resource constraints. To overcome this problem, this paper presents a Blockchain-based security mechanism that enables secure authorized access to smart city resources. The presented mechanism comprises the ACE (Authentication and Authorization for Constrained Environments) framework-based authorization Blockchain and the OSCAR (Object Security Architecture for the Internet of Things) object security model. The Blockchain lays out a flexible and trustless authorization mechanism, while OSCAR makes use of a public ledger to structure multicast groups for authorized clients. Moreover, a meteor-based application is developed to provide a user-friendly interface for heterogeneous technologies belonging to the smart city. The users would be able to interact with and control their smart city resources such as traffic lights, smart electric meters, surveillance cameras, etc., through this application. To evaluate the performance and feasibility of the proposed mechanism, the authorization Blockchain is implemented on top of the Ethereum network. The authentication mechanism is developed in the node.js server and a smart city is simulated with the help of Raspberry Pi B+. Furthermore, mocha and chai frameworks are used to assess the performance of the system. Experimental results reveal that the authentication response time is less than 100 ms even if the average hand-shaking time increases with the number of clients.

Visually impaired persons (VIPs) comprise a significant portion of the population, and they are present around the globe and in every part of the world. In recent times, technology proved its presence in every domain, and innovative devices assist humans in their daily lives. In this work, a smart and intelligent system is designed for VIPs to assist mobility and ensure their safety. The proposed system provides navigation in real-time using an automated voice. Though VIPs wouldn't be able to see objects in their surroundings, they can sense and visualize the roaming environment. Moreover, a web-based application is developed to ensure their safety. The user of this application can turn the on-demand function for sharing his/her location with the family while compromising privacy. Through this application, the family members of VIPs would be able to track their movement (get location and snapshots) while being at their homes. Hence, the device allows VIPs to visualize the environment and ensure their security. Such a comprehensive device was a missing link in the existing literature. The application uses MobileNet architecture due to its low computational complexity to run on low-power end devices. To assess the efficacy of the proposed system, six pilot studies have been performed that reflected satisfactory results. For object detection and recognition, a deep Convolution Neural Network (CNN) model is employed with an accuracy of 83.3%, whereas the dataset contains more than 1000 categories. Moreover, a score-based quantitative comparative analysis is performed using the supported features of devices. It is found that the proposed system has outperformed the existing devices having a total score of 9.1/10, which is 8% higher than the second-best.