Computer Science Review最新文献

Electricity is one of the mandatory commodities for mankind today. To address challenges and issues in the transmission of electricity through the traditional grid, the concepts of smart grids and demand response have been developed. In such systems, a large amount of data is generated daily from various sources such as power generation (e.g., wind turbines), transmission and distribution (microgrids and fault detectors), load management (smart meters and smart electric appliances). Thanks to recent advancements in big data and computing technologies, Deep Learning (DL) can be leveraged to learn the patterns from the generated data and predict the demand for electricity and peak hours. Motivated by the advantages of deep learning in smart grids, this paper sets to provide a comprehensive survey on the application of DL for intelligent smart grids and demand response. Firstly, we present the fundamental of DL, smart grids, demand response, and the motivation behind the use of DL. Secondly, we review the state-of-the-art applications of DL in smart grids and demand response, including electric load forecasting, state estimation, energy theft detection, energy sharing and trading. Furthermore, we illustrate the practicality of DL via various use cases and projects. Finally, we highlight the challenges presented in existing research works and highlight important issues and potential directions in the use of DL for smart grids and demand response.

The music domain is among the most important ones for adopting recommender systems technology. In contrast to most other recommendation domains, which predominantly rely on collaborative filtering (CF) techniques, music recommenders have traditionally embraced content-based (CB) approaches. In the past years, music recommendation models that leverage collaborative and content data – which we refer to as content-driven models – have been replacing pure CF or CB models. In this survey, we review 55 articles on content-driven music recommendation. Based on a thorough literature analysis, we first propose an onion model comprising five layers, each of which corresponds to a category of music content we identified: signal, embedded metadata, expert-generated content, user-generated content, and derivative content. We provide a detailed characterization of each category along several dimensions. Second, we identify six overarching challenges, according to which we organize our main discussion: increasing recommendation diversity and novelty, providing transparency and explanations, accomplishing context-awareness, recommending sequences of music, improving scalability and efficiency, and alleviating cold start. Each article addresses one or more of these challenges and is categorized according to the content layers of our onion model, the article’s goal(s), and main methodological choices. Furthermore, articles are discussed in temporal order to shed light on the evolution of content-driven music recommendation strategies. Finally, we provide our personal selection of the persisting grand challenges which are still waiting to be solved in future research endeavors.

Quantum physics has changed the way we understand our environment, and one of its branches, quantum mechanics, has demonstrated accurate and consistent theoretical results. Quantum computing is the process of performing calculations using quantum mechanics. This field studies the quantum behavior of certain subatomic particles (photons, electrons, etc.) for subsequent use in performing calculations, as well as for large-scale information processing. These advantages are achieved through the use of quantum features, such as entanglement or superposition. These capabilities can give quantum computers an advantage in terms of computational time and cost over classical computers. Nowadays, scientific challenges are impossible to perform by classical computation due to computational complexity (more bytes than atoms in the observable universe) or the time it would take (thousands of years), and quantum computation is the only known answer. However, current quantum devices do not have yet the necessary qubits and are not fault-tolerant enough to achieve these goals. Nonetheless, there are other fields like machine learning, finance, or chemistry where quantum computation could be useful with current quantum devices. This manuscript aims to present a review of the literature published between 2017 and 2023 to identify, analyze, and classify the different types of algorithms used in quantum machine learning and their applications. The methodology follows the guidelines related to Systematic Literature Review methods, such as the one proposed by Kitchenham and other authors in the software engineering field. Consequently, this study identified 94 articles that used quantum machine learning techniques and algorithms and shows their implementation using computational quantum circuits or ansatzs. The main types of found algorithms are quantum implementations of classical machine learning algorithms, such as support vector machines or the k-nearest neighbor model, and classical deep learning algorithms, like quantum neural networks. One of the most relevant applications in the machine learning field is image classification. Many articles, especially within the classification, try to solve problems currently answered by classical machine learning but using quantum devices and algorithms. Even though results are promising, quantum machine learning is far from achieving its full potential. An improvement in quantum hardware is required for this potential to be achieved since the existing quantum computers lack enough quality, speed, and scale to allow quantum computing to achieve its full potential.

The growing technology of the fifth generation (5G) of mobile telecommunications has led to the special attention of cloud service providers (CSPs) to mobile cloud computing (MCC). Due to the limitations in processing power, storage space and energy capacity of mobile devices, cloud resources can be moved to the edge of the network to improve the quality of service (QoS). Server placement is a crucial emerging problem in both typical and edge types of MCC, different proposed methods of which are reviewed and evaluated in this paper. Proper placement of servers leads to more efficient utilization of these servers, reduces their response time and optimizes their energy consumption. A variety of techniques and approaches, including machine learning-based techniques, evolutionary models, optimization algorithms, heuristics and meta-heuristics have been employed by different server placement methods of the literature to find the optimal deployment map of servers. This paper provides a comprehensive analysis of these server placement methods in edge computing, fog computing and cloudlet, investigates their various aspects, dimensions and objectives, and evaluates their strengths and weaknesses. Furthermore, open challenges for server placement in MCC are provided, and future research directions are also explained and discussed.

As a new type of aerial robotics, drones are easy to use and inexpensive, which has facilitated their acquisition by individuals and organizations. This unequivocal and widespread presence of amateur drones may cause many dangers, such as privacy breaches by reaching sensitive locations of authorities and individuals. In this paper, we summarize the performance-affecting factors and major obstacles to drone use and provide a brief background of deep learning. Then, we summarize the types of UAVs and the related unethical behaviors, safety, privacy, and cybersecurity concerns. Then, we present a comprehensive literature review of current drone detection methods based on deep learning. This area of research has arisen in the last two decades because of the rapid advancement of commercial and recreational drones and their combined risk to the safety of airspace. Various deep learning algorithms and their frameworks with respect to the techniques used to detect drones and their areas of applications are also discussed. Drone detection techniques are classified into four categories: visual, radar, acoustics, and radio frequency-based approaches. The findings of this study prove that deep learning-based detection and classification of drones looks promising despite several challenges. Finally, we provide some recommendations to meet future expectations.

Nash equilibrium is one of the most influential solution concepts in game theory. With the development of computer science and artificial intelligence, there is an increasing demand on Nash equilibrium computation, especially for Internet economics and multi-agent learning. This paper reviews various algorithms computing the Nash equilibrium and its approximation solutions in finite normal-form games from both theoretical and empirical perspectives. For the theoretical part, we classify algorithms in the literature and present basic ideas on algorithm design and analysis. For the empirical part, we present a comprehensive comparison on the algorithms in the literature over different kinds of games. Based on these results, we provide practical suggestions on implementations and uses of these algorithms. Finally, we present a series of open problems from both theoretical and practical considerations.

In recent years, the number of crimes with weapons has grown on a large scale worldwide, mainly in locations where enforcement is lacking or possessing weapons is legal. It is necessary to combat this type of criminal activity to identify criminal behavior early and allow police and law enforcement agencies immediate action. Despite the human visual structure being highly evolved and able to process images quickly and accurately if an individual watches something very similar for a long time, there is a possibility of slowness and lack of attention. In addition, large surveillance systems with numerous equipment require a surveillance team, which increases the cost of operation. There are several solutions for automatic weapon detection based on computer vision; however, these have limited performance in challenging contexts. A systematic review of the current literature on deep learning-based weapon detection was conducted to identify the methods used, the main characteristics of the existing datasets, and the main problems in the area of automatic weapon detection. The most used models were the Faster R-CNN and the YOLO architecture. The use of realistic images and synthetic data showed improved performance. Several challenges were identified in weapon detection, such as poor lighting conditions and the difficulty of small weapon detection, the last being the most prominent. Finally, some future directions are outlined with a special focus on small weapon detection.

The exploding usage of physical object properties has greatly facilitated real-time applications such as robotics to perceive exactly as it appears in existence. Changes in the nature and properties of diverse real-time systems are associated with physical properties modification due to environmental factors. These physics-based object properties features attract the researchers’ attention while developing solutions to real-life problems. But, the detection and prediction of physical properties change are very diverse, covering many physics laws and object properties (material, shape, gravitational force, color, state change) which append complexity to these tasks. Instead of well-understood physics laws, elucidating physics laws requires substantial manual modeling with the help of standardized equations and associated factors. To adopt these physical laws to get instinctive and effective outcomes, researchers started applying computational models to learn changing property behavior as a substitute for using handcrafted and equipment-generated variable states. If physical properties detection challenges are not anticipated and required measures are not precluded, the upcoming computational model-driven physical object changing will not be able to serve appropriately. Therefore, this survey paper is drafted to demonstrate comprehensive theoretical and empirical studies of physical object properties detection and prediction. Furthermore, a generic paradigm is proposed to work in this direction along with characterization parameters of numerous physical object properties. A brief summarization of applicable machine learning, deep learning, and metaheuristic approaches is presented. An extensive list of released and openly available datasets for varying objects and parameters rendered for researchers. Additionally, performance measures regarding computational techniques for physical properties discovery and detection for quantitative evaluation of outcomes are also entailed. Finally, a few open research issues that need to be explored in the future are specified.

The emergence of ubiquitous computing and different disruptive technologies caused magnificent development in information and communication technology. Likewise, cybercriminals are also carefully considering different newer ways of attacks. Protecting the confidentiality, integrity, and authentication of sensitive information is the day’s major challenge. Secret sharing is a method that allows a trusted authority (the dealer) to distribute a secret or a number of secrets among some target participants with the intention that certain predetermined groups of participants can collaborate to recover the secret or secrets. Any other group formed by the participants cannot do so. Threshold secret sharing (TSS) is a particular form of secret sharing. It permits any group consisting of at least a specific number (called the threshold) of participants to reconstruct the secret or secrets. However, any group with fewer than the specified number of participants is forbidden to do so. It provides tolerance against single point of failure (SPOF), which has attracted a large number of researchers to contribute in this field. It has the potential to be implemented in numerous practical and secure applications. In this paper, we present a comprehensive survey of a variety of existing threshold secret sharing schemes. We have identified various aspects of developing secure and efficient secret sharing schemes. We have also highlighted some of the applications based on secret sharing. Finally, the open challenges and future research directions in the field of secret sharing are identified and discussed.

The continuous increase of IoT applications leads to a vast amount of data that needs to be transmitted, stored, and processed. Many IoT applications rely on the Cloud infrastructure to handle these specific application demands. However, the integration of IoT and Cloud poses challenges such as network delays, throughput, energy consumption, reliability, etc. Therefore, a new computing concept is required to support emerging IoT applications. These new concepts include fog computing, edge computing, mobile edge computing, mobile cloud computing, and cloudlets. They use various approaches to distribute resources, processes, and services among IoT system architecture layers. The challenge is to decide which offloading system is the best for a specific use case that emphasizes the IoT system modeling issue. In this paper, a model for the formal description of IoT systems is presented. In addition, an analytical evaluation method was proposed to design these systems using the corresponding architecture, technologies, protocols, and integration model to optimize performance. The proposed approach facilitates and simplifies the selection of the corresponding model for the system architecture. This approach enables an efficient method for performance optimization based on offloading processes (load balancing). Also, this paper provides some insights into specific emerging issues and ideas to be addressed by future research.