Recurrent neural networks (RNNs) are an excellent fit for regression problems where sequential data are the norm since their recurrent internal structure can analyse and process data for long. However, RNNs are prone to the phenomenal vanishing gradient problem (VGP) that causes the network to stop learning and generate poor prediction accuracy, especially in long-term dependencies. Originally, gated units such as long short-term memory (LSTM) and gated recurrent unit (GRU) were created to address this problem. However, VGP was and still is an unsolved problem, even in gated units. This problem occurs during the backpropagation process when the recurrent network weights tend to vanishingly reduce and hinder the network from learning the correlation between temporally distant events (long-term dependencies), that results in slow or no network convergence. This study aims to provide an empirical analysis of LSTM networks with an emphasis on inefficiency in long-term dependencies convergence because of VGP. Case studies on NASA’s turbofan engine degradation are examined and empirically analysed.

Cloud computing is a rapidly growing paradigm which has evolved from having a monolithic to microservices architecture. The importance of cloud data centers has expanded dramatically in the previous decade, and they are now regarded as the backbone of the modern economy. Cloud-based microservices architecture is incorporated by firms such as Netflix, Twitter, eBay, Amazon, Hailo, Groupon, and Zalando. Such cloud computing arrangements deal with the parallel deployment of data-intensive workloads in real time. Moreover, commonly utilized cloud services such as the web and email require continuous operation without interruption. For that purpose, cloud service providers must optimize resource management, efficient energy usage, and carbon footprint reduction. This study presents a conceptual framework to manage the high amount of microservice execution while reducing response time, energy consumption, and execution costs. The proposed framework suggests four key agent services: (1) intelligent partitioning: responsible for microservice classification; (2) dynamic allocation: used for pre-execution distribution of microservices among containers and then makes decisions for dynamic allocation of microservices at runtime; (3) resource optimization: in charge of shifting workloads and ensuring optimal resource use; (4) mutation actions: these are based on procedures that will mutate the microservices based on cloud data center workloads. The suggested framework was partially evaluated using a custom-built simulation environment, which demonstrated its efficiency and potential for implementation in a cloud computing context. The findings show that the engrossment of suggested services can lead to a reduced number of network calls, lower energy consumption, and relatively reduced carbon dioxide emissions.

Secure communication is the top concern of the Internet of Vehicles (IoV). The trust between nodes can have a considerable impact on ensuring IoV security. Therefore, the trustworthiness of a received message must be evaluated before acting upon it. A malicious node can broadcast bogus events to obtain network control. False reports and malicious vehicles render the network unreliable during emergencies. In this study, a unique trust framework is presented that considers most of the aspects of trust in IoV to accurately identify malicious nodes and events. Previous studies have proposed some trust models for VANETs, which have many deficiencies in serving IoV. In particular, they lack dynamism and practical implementations. All the existing models have two things in common, first they work on fixed parameters, and second, they use static scenarios. In contrast, the proposed framework is based on a context-awareness cognitive approach with artificial intelligence (AI) properties. The framework cognitively learns the environment from the received report and creates a context around an event. In addition to trust management (TM), the proposed framework offers a novel process for detecting and screening malicious nodes using anomaly outliers. The performance of the framework was examined using an experimental simulation. The proposed framework was compared with top benchmarks in the field. The results show inclining performance indicators. The proposed trust-management framework has the potential to serve as a component of IoV security.