The leader node in a distributedcomputing system is responsible to establish coordination among all other nodes that are situated apart geographically. Selection of a suitable leader is one of the major and challengi...
详细信息
The leader node in a distributedcomputing system is responsible to establish coordination among all other nodes that are situated apart geographically. Selection of a suitable leader is one of the major and challenging problems. In this paper, a novel leader election algorithm is proposed based on resources of each node in a ring network. All the nodes compute resource strength values by considering available resources like CPU, memory capacity, and residual energy. A node with the highest resource strength over the network is elected as the leader. The proposed algorithm has also considered sudden failure of the nodes particularly when it is the leader node. Moreover, addition of new nodes is also considered. The proposed algorithm shows improvement on message complexity over the network and resource-based priority generation, which helps in efficient election of the leader. To validate, the proposed algorithm is extensively simulated as well as real-life hardware experiment is also done. In the experiment, 2 cases are considered with different weight of the resources, and consequent effects are shown. The results are also compared with the existing algorithms, and it is observed that the proposed work comparably performs better that the existing related algorithms.
The paper presents an analysis of the use of optimization algorithms in parallel solutions and distributed computing systems. The primary goal is to use evolutionary algorithms and their implementation into parallel c...
详细信息
The paper presents an analysis of the use of optimization algorithms in parallel solutions and distributed computing systems. The primary goal is to use evolutionary algorithms and their implementation into parallel calculations. Parallelization of computational algorithms is suitable for the following cases computational models with a large number of design variables or cases where the objective function evaluation is time consuming (FE analysis). As the software platform for application of distributed optimization algorithms is using MATLAB and BOINC software package.
The method of evaluating the effectiveness of the economic parameters for the implementation at industrial enterprises of "Peer-to-Peer distributed Storage and Processing of Protected Information in Corporate Net...
详细信息
ISBN:
(纸本)9781728167015
The method of evaluating the effectiveness of the economic parameters for the implementation at industrial enterprises of "Peer-to-Peer distributed Storage and Processing of Protected Information in Corporate Networks," designed primarily to protect distributedcomputing and isolate private networks without limiting the possibilities of useful interaction in the framework of the technologies Industry 5.0 and Super Smart Society is proposed. Within the framework of these concepts, there is a gradual integration of production, logistics, sales, and information systems into a global information exchange system. The proposed system provides for effective counteraction to the threats using cryptography, protection against unauthorized access using biometrics, and an innovative protocol for information interaction and control of topology based on distributed ledger technologies, it optimizes the use of computing power and provides tools for distributedcomputing without changing the logic of existing software.
Although the principles of real-time collaborative editing have been explored since the eighties, team collaboration software facilitating the completion of tasks as a group continues to be a very hot research topic. ...
详细信息
ISBN:
(纸本)9781538646403
Although the principles of real-time collaborative editing have been explored since the eighties, team collaboration software facilitating the completion of tasks as a group continues to be a very hot research topic. A series of theoretical and practical results obtained by the research and industrial communities originated in the theory of distributedcomputing. They were devised for managing the concurrent nature of user actions and for maintaining the consistency of data as changes are introduced randomly, by multiple users and in real-time. As such, centralized collaborative editing servers were designed to allow users to work in parallel on a document from a typical web browser. In order to maintain the consistency of the content being modified at different sites in different orders, Operational Transformation (OT) mechanisms are at the core of collaboration servers enabling web-based co-editing. However, as expected of modern web application deployments, a centralized OT algorithm is required that must also exhibit properties such as scalability and reliability. In this paper, the processes involved in the client-server interactions of OT are modeled as real-time systems using Finite State Machine (FSM) theory. The consistency of the data is controlled by formal groups of FSMs. Hierarchical FSMs are used to define and simulate the real-time behavior of client and server components when processing and transforming changes initiated by users. The FSM-based OT implementation is tested using random inputs and the approach is shown to be helpful for organizing and managing the complex distributed aspects of such algorithms.
This work is devoted to the problem of detecting and processing faults of computing nodes during execution of parallel programs on distributed computing systems. The fault tolerance tools of PBS/TORQUE are considered....
详细信息
ISBN:
(纸本)9781728129877
This work is devoted to the problem of detecting and processing faults of computing nodes during execution of parallel programs on distributed computing systems. The fault tolerance tools of PBS/TORQUE are considered. The functional model for faults handling optimization are proposed.
The availability of Healthcare Data Center (HDC) from network or machine failures has led to the appearance of new approaches to self healing data centres. Modern data centers looking to ensure the reliability of Dist...
详细信息
ISBN:
(纸本)9781538621073;9781538621066
The availability of Healthcare Data Center (HDC) from network or machine failures has led to the appearance of new approaches to self healing data centres. Modern data centers looking to ensure the reliability of distributed computing systems (DCS) in the most critical era. Also, modern HDC need to provide highest reliability infrastructure for all distribution critical jobs and data sources including electronic health records, electronic medical records and IoT. Redundancy is the traditional technique that improves the reliability of DCS among HDC. Inherently, redundancy is a very expensive approach in terms of hardware and software resources. In this context, this paper presents an Enhanced MapReduce Agent Mobility (EMRAM) framework to maximize reliability of DCSs among HDC without redundancy. The proposed framework migrating tasks to the available machines in the in HDC when identifying unhampered machine failure. Some practical experiments are performed to evaluate the proposed (EMRAM) framework by considering different scenarios. The obtained results are compared with the most recent frameworks showing a relatively better performance and higher reliability than other recent frameworks.
Effective scheduling of the tasks of a distributed application is one of the key factors in achieving improved performance. It results in an adequate utilization of the underlying resources and also reduces the total ...
详细信息
Effective scheduling of the tasks of a distributed application is one of the key factors in achieving improved performance. It results in an adequate utilization of the underlying resources and also reduces the total execution time of the application. Generating an optimal schedule for a distributed application is not a trivial task as it exists in the class of NP-complete problems. In this paper, a novel strategy called incremental subgraph earliest finish time (INCSEFT) is proposed. It is aimed at scheduling tasks on heterogeneous systems. It incorporates the use of a subgraph that grows incrementally by adding critical paths. At each step, the scheduling strategy attempts to minimize the schedule length. Considering a large set of nodes at an instance makes this approach perform better than other scheduling strategies used for heterogeneous systems. The experiments performed with several graphs show that the INCSEFT strategy produces significant improvement over the well-known HEFT, LOOKAHEAD and CEFT strategies used for scheduling heterogeneous systems.
Modern distribution system are expected to provide new features such as taking advantage of Cyber-Physical systems (CPS) - new equipment and devices embedded with sensors, network communication, and computational inte...
详细信息
ISBN:
(纸本)9781509046010
Modern distribution system are expected to provide new features such as taking advantage of Cyber-Physical systems (CPS) - new equipment and devices embedded with sensors, network communication, and computational intelligence techniques to provide increased system performance and power quality. Among the performance improvement, the reduction of electrical losses is an important quality factor which is associated with energy efficiency. This paper presents a method based on Multi-agent systems (MAS) that manages topology changes by switching operations to improve the system performance in dynamic scenario, where the power demand varies throughout the day. Experiments were performed allocating three different load consumer profiles (residential, commercial, and industrial) in two test systems with 12-bus and 16-bus, creating several scenarios. The agents were deployed in a set of small-sized single-board computers with low computational power to mimic CPS. The simulations has shown the success of the method on managing the decision making among different agents to provide the joint effort to manage the loss reduction on the network.
This article considers up-to-date problems of parallel computingsystems development in Russia. The study describes an engineering of an interface for organization of high-performance system with programmable structur...
详细信息
ISBN:
(纸本)9781509040698
This article considers up-to-date problems of parallel computingsystems development in Russia. The study describes an engineering of an interface for organization of high-performance system with programmable structure and for its partition to specific subsystems focused on applied tasks. The solution is based on the concept of homogeneous computingsystems (HCS) relying on the model of collective of calculators. HCS concept was formulated in the Institute of Mathematics of the USSR under the leadership of E.V. Evreinov in 1962 [1]. The approach is particularly relevant in terms of growing popularity of information networks which combining numerous personal computers (PCs). It's obviously necessary to develop software tools to integrate disparate calculators into the system and identify the required number of sub-systems for specific problem solving. An interference into the hardware and software parts of the calculators must be avoided during the software development process. Programmable structure of the computer system is provided at the stage of unification of calculators into software subsystem interface called an agent of DCS PS. The experience of the development of the Interface organization of distributed computing systems with programmable structure and allocation of subsystems for specific applications. The software described in the present document has been tested while an experimental organization of the DCS PS at the department of the computer engineering of NSTU.
As a mobile operating system framework, Android plays a significant role in supporting mobile apps. However, current Android application model is not efficient by using current two common approaches, including Activit...
详细信息
As a mobile operating system framework, Android plays a significant role in supporting mobile apps. However, current Android application model is not efficient by using current two common approaches, including Activity+XML Layout Files (AXLF) and HTML+WebKit (HWK) models. In this paper, we propose a novel middleware service solution that overcomes the drawbacks with using the pre-cache approach, PrecAche Technology of Android System (PATAS). The proposed method uses HTML to design the application interface and separately store the Page Framework (PF) and Page Data (PD). We create a new middleware of web pages, Version Flags, to indicate whether PF and PD are expired. Our experimental results represent that the proposed approach can improve the execution efficiency as well as reduce the networking costs, which can be broadly used in cloud-based distributedsystems. (C) 2015 Elsevier B.V. All rights reserved.
暂无评论