Blockchain is a modern technology that has revolutionized the way society interacts and trades. It could be defined as a chain of blocks that stores information with digital signatures in a distributed and decentralized network. This technique was first adopted for the creation of digital cryptocurrencies, such as Bitcoin and Ethereum. However, research and industrial studies have recently focused on the opportunities that blockchain provides in various other application domains to take advantage of the main features of this technology, such as: decentralization, persistency, anonymity, and auditability. This paper reviews the use of blockchain in several interesting fields, namely: finance, healthcare, information systems, wireless networks, Internet of Things, smart grids, governmental services, and military/defense. In addition, our paper identifies the challenges to overcome, to guarantee better use of this technology.

Long Short-Term Memory (LSTM) networks have revolutionized the field of deep learning, particularly in applications that require the modeling of sequential data. Originally designed to overcome the limitations of traditional recurrent neural networks (RNNs), LSTMs effectively capture long-range dependencies in sequences, making them suitable for a wide array of tasks. This survey aims to provide a comprehensive overview of LSTM architectures, detailing their unique components, such as cell states and gating mechanisms, which facilitate the retention and modulation of information over time. We delve into the various applications of LSTMs across multiple domains, including the following: natural language processing (NLP), where they are employed for language modeling, machine translation, and sentiment analysis; time series analysis, where they play a critical role in forecasting tasks; and speech recognition, significantly enhancing the accuracy of automated systems. By examining these applications, we illustrate the versatility and robustness of LSTMs in handling complex data types. Additionally, we explore several notable variants and improvements of the standard LSTM architecture, such as Bidirectional LSTMs, which enhance context understanding, and Stacked LSTMs, which increase model capacity. We also discuss the integration of Attention Mechanisms with LSTMs, which have further advanced their performance in various tasks. Despite their strengths, LSTMs face several challenges, including high Computational Complexity, extensive Data Requirements, and difficulties in training, which can hinder their practical implementation. This survey addresses these limitations and provides insights into ongoing research aimed at mitigating these issues. In conclusion, we highlight recent advances in LSTM research and propose potential future directions that could lead to enhanced performance and broader applicability of LSTM networks. This survey serves as a foundational resource for researchers and practitioners seeking to understand the current landscape of LSTM technology and its future trajectory.

Human behavior modeling in smart environments is a growing research area treating several challenges related to ubiquitous computing, pattern recognition, and ambient assisted living. Thanks to recent progress in sensing devices, it is now possible to design computational models able of accurate detection of residents’ activities and daily routines. For this goal, we introduce in this paper a deep learning-based framework for activity recognition in smart homes. This framework proposes a detailed methodology for data preprocessing, feature mining, and deep learning techniques application. The novel framework was designed to ensure a deep exploration of the feature space since three main approaches are tested, namely, the all-features approach, the selection approach, and the reduction approach. Besides, the framework proposes the evaluation and the comparison of several well-chosen deep learning techniques such as autoencoder, recurrent neural networks (RNN), and some of their derivatives models. Concretely, the framework was applied on the “Orange4Home” dataset which represents a recent dataset specially designed for smart homes research. Our main findings show that the best approach for efficient classification is the selection approach. Furthermore, our overall results outperformed baseline models based on random forest classifiers and the principal component analysis technique, especially the results of our RNN-based model for the all-features approach and the results of our autoencoder-based model for the feature reduction approach.

The usage of wearable gadgets is growing in the cloud-based health monitoring systems. The signal compression, computational and power efficiencies play an imperative part in this scenario. In this context, we propose an efficient method for the diagnosis of cardiovascular diseases based on electrocardiogram (ECG) signals. The method combines multirate processing, wavelet decomposition and frequency content-based subband coefficient selection and machine learning techniques. Multirate processing and features selection is used to reduce the amount of information processed thus reducing the computational complexity of the proposed system relative to the equivalent fixed-rate solutions. Frequency content-dependent subband coefficient selection enhances the compression gain and reduces the transmission activity and computational cost of the post cloud-based classification. We have used MIT-BIH dataset for our experiments. To avoid overfitting and biasness, the performance of considered classifiers is studied by using five-fold cross validation (5CV) and a novel proposed partial blind protocol. The designed method achieves more than 12-fold computational gain while assuring an appropriate signal reconstruction. The compression gain is 13 times compared to fixed-rate counterparts and the highest classification accuracies are 97.06% and 92.08% for the 5CV and partial blind cases, respectively. Results suggest the feasibility of detecting cardiac arrhythmias using the proposed approach.

The state of roads may sometimes be difficult to perceive due to intense climate conditions, absence of road signs, or simply human inattention, which may be harmful to both vehicles and drivers. The automatic monitoring of the road states represents a promising solution to warn drivers about the status of a road in order to protect them from injuries or accidents. In this paper, we present a novel application for data collection regarding road states. Our application entitled “Road Scanner” allows onboard users to tag four types of segments in roads: smooth, bumps, potholes, and others. For each tagged segment the application records multimodal data using the embedded sensors of a smartphone. The collected data concerns mainly vehicle accelerations, angular rotations, and geographical positions recorded by the accelerometer, the gyroscope, and the GPS sensor, respectively, of a user phone. Moreover, a medium-size dataset was built and machine learning models were applied to detect the right label for the road segment. Overall, the results were very promising since the SVM classifier (Support Vector Machines) has recorded an accuracy rate of 88.05%.

Adversarial attacks involve manipulating data to trick Artificial Intelligence (AI) models, making false predictions or classifications or even disrupting the normal functions of the smart grid. This can be done by providing the wrong information to the models, hence producing wrong predictions and therefore leading to instabilities, power imbalances and overall operational failure. Thus, employing an AI model is vital for controlling energy usage, identifying and even predicting equipment failures, and accurately determining power availability. Thus, their input dependency does not enable them to withstand cyber-attacks, which compromises the stability of the grid. These attacks affect energy control, cause losses and intrude into the critical infrastructure part of the smart grid, hence the call for enhanced smart grid protection. This study first generates a novel adversarial attack dataset on smart grids with three attacks: adversarial perturbation, backdoor injection, DOS attacks, and one benign class. Next, it provides a fine-tuned Deep Neural Network (DNN) model to improve resistance against adversarial attacks on smart grids significantly. Results from the various Machine Learning (ML) and DNN algorithms showed accuracy varying between 29.10% to 73.9%, with DNN recording the highest levels of accuracy. This shows how the approach can be utilised in the prediction of such attacks and how the grid can be protected and secured.