Game Theoretical Data Replication Techniques For Large Scale Autonomous Distributed Computing Systems
Download Game Theoretical Data Replication Techniques For Large Scale Autonomous Distributed Computing Systems PDF/ePub or read online books in Mobi eBooks. Click Download or Read Online button to get Game Theoretical Data Replication Techniques For Large Scale Autonomous Distributed Computing Systems book now. This website allows unlimited access to, at the time of writing, more than 1.5 million titles, including hundreds of thousands of titles in various foreign languages.
Game Theoretical Data Replication Techniques for Large-scale Autonomous Distributed Computing Systems
Data replication in geographically dispersed servers is an essential technique for reducing the user perceived access time in large-scale distributed computing systems. A majority of the conventional replica placement techniques lack scalability and solution quality. To counteract such issues, this thesis proposes a game theoretical replica placement framework, in which autonomous agents compete for the allocation or reallocation of replicas onto their representative servers in a self-managed fashion. Naturally, each agent's goal is to maximize its own benefit. However, the framework is designed to suppress individualism and to ensure system-wide optimization. Using this framework as an environment, several cooperative and non-cooperative low-complexity, flexible, and scalable game theoretical replica placement techniques are proposed, analytically investigated, and experimentally evaluated. Each of these techniques supports different game theoretical (pareto-optimality, catering to agents' interests, deliberate discrimination of allocation, budget balanced, pure Nash equilibrium, and Nash equilibrium) and system (link distance, congestion control, minimization of communication cost, and memory optimization) related properties. Using a detailed test-bed involving eighty various network topologies and two real-world access logs, each game theoretical technique is also extensively compared with conventional replica placement techniques, such as, greedy heuristics, branch-and-bound techniques and genetic algorithms. The experimental study confirms that in each case the proposed techniques outperform other conventional methods. The results can be summarized in four ways: (1) The number of replicas in a system self-adjusts to reflect the ratio of the number of reads versus writes access; (2) Performance is improved by replicating objects to the servers based on the locality of reference; (3) Replica allocations are made in a fast algorithmic turn-around time; (4) The complexity of the data replication problem is decreased by multifold.
Handbook of Parallel Computing
The ability of parallel computing to process large data sets and handle time-consuming operations has resulted in unprecedented advances in biological and scientific computing, modeling, and simulations. Exploring these recent developments, the Handbook of Parallel Computing: Models, Algorithms, and Applications provides comprehensive coverage on a
Vehicle Computing
Over the past century, vehicles have predominantly functioned as a means of transportation. However, as vehicular computation and communication capacities continue to expand, it is anticipated that upcoming connected vehicle (CVs) will not only serve their conventional transport functions but also act as versatile mobile computing platforms. This book presents the concept of Vehicle Computing, encompassing five primary functionalities of CVs: computation, communication, energy management, sensing, and data storage. It proposes a potential business model and explores the challenges and opportunities associated with these domains. Vehicle Computing serves as an important resource for the research community and practitioners in the field of edge computing and cyber physical system, capturing the essence of a rapidly changing industry, addressing the challenges and opportunities associated with connected vehicles (including software-defined vehicles, autonomous vehicles, electric vehicles), machine learning, communication, sensing, data storage, energy management, and computer systems. It synthesizes the latest research and real-world applications, offering valuable insights to both academia and industry professionals. Vehicle Computing covers topics such as: The fundamentals of Vehicle Computing, including its historical context and key components. Advanced communication and networking technologies for connected vehicles. Sensing and data acquisition techniques, including edge and cloud computing integration. Energy management and storage, focusing on electric vehicle infrastructure and vehicle-to-grid. Data storage and processing strategies for vehicular environments. Business models, opportunities, and challenges associated with Vehicle Computing. Real-world applications and case studies, highlighting best practices and future trends.