مجله محاسبات و سامانه های توزیع شده
پیاپی 2 (پاییز و زمستان 1397)

The internet has been developed for the first time in 1999 by Kevin Ashton, based on statistics published by research and research institute IDC as one of the highest - earning projects since 2006. One of the most important issues on the internet research indicates that the architecture of the internet has a three - layer model.This study presents a five - layer model for the internet by analyzing the proposed model, and by analyzing and examining the proposed model, the proposed model will provide a secure authentication service relative to the three layer models in order to improve the resistance against external intervention as well as Distributed confidentiality. The results of this study show that the blockchain - based method for increasing privacy security and capability of control will be a practical solution to the issue of internet security.

In the information tecknology era when organizations are growing so fast and business processes are becoming more complicated, their management and controlling would become more important than before, then improving business processes as well as business intelligence is unavoidable. Meanwhile, business processes management would be accomplished as a strategy to increase efficiency, effectiveness, and reliability of organizations' business processes happening through compiling and modeling them. BI aims to analyze processes and making exact and intelligent decisions plays a vital role in effective decision-making in the commercial field in order to improve business performance. Changing business concepts to model, we can choose the modeling business process as the first priority and an accepted managerial tool. So, we exploit ITLI framework in this research to exploit standard architecture in modelling business processes.

We live in a world where consumer products, vehicles, industrial parts and tools, sensors and other everyday objects are combined with powerful Internet connectivity and powerful analysis systems. Through the internet of things, we are producing a large amount of data. The internet of things is a new concept and computing that makes our world smarter, with the help of this technology, all objects will always be able to interact with each other and generate a lot of data from their own situation. With the growth of the speed of production and the unprecedented level of data, cloud-based infrastructure alone does not meet the needs of this technology, and it should be borne in mind that the functionality of the Internet of objects is reliant on the cloud, which can lead to inefficiencies. Latency and bandwidth are two important factors in cloud computing inefficiency, due to the large amount of data and the high processing latency, the cloud can't respond promptly to incidents and events. The lack of appropriate bandwidth also the network of internet objects damages, the devices produce a lot of data in the network, which, if it is not possible to send all these data to the cloud for processing, can destroy important information for decision making. Therefore, the internet of things needs a solution to address this challenge. For this purpose, false computing or edge computing has been focused on removing these constraints. In this type of computing, which is locally and decentralized, data processing is done near the data production site. Combining its cloud computing solution can meet the needs of the internet technology of objects. Despite the increasing use of cloud computing, there are still unresolved issues due to inherent problems of cloud computing such as unreliable latency, lack of mobility support and location. Information is there. Therefore, we need a computational paradigm that allows us to compute the data before sending it to the super-batch, which allows us to work on the data before it is lost and lost. May's computing is a paradigm that expands cloud computing by sending services to the edge of the network, and dew computing is an organized computerized computing pattern in the era of cloud computing. Since the launch of the internet work, it has been continuously serving as the main pillar of the existing aspects of cloud, fog, and edge computing. In this paper, we describe the various properties of dew computing and describe some of its applications, as well as their parametric comparison with cloud computing paradigms.

Load imbalance, multiple outages, high voltage drop and high losses are one of the major problems in distribution networks. One of the solutions to solving these issues today is to rearrange the distribution networks. The rearrangement of distribution systems can have different goals. Most researchers have considered reducing losses, balancing the load of feeders, and improving the voltage profile as the main objectives of the rearrangement. In the last decade, many methods have been used for rearranging, which have been remarkably used the heuristic and meta-heuristic methods. In this paper, an optimal rearrangement of the distribution system is carried out using the Particle Swarm Optimization (PSO) algorithm, inspired by the collective movement of birds. The results of simulation in MATLAB indicate the effectiveness of the proposed approach to reduce losses, improves the voltage profile and load balancing.

In this paper, an adaptive unit is designed for distance relay using the Artificial Neural Network (ANN) to study the effect of power oscillation on distances relays in the presence of distributed generation sources. The proposed system, along with the relay constant setting unit, uses the local information to set the trip limit for the relay. In this case, the adjustment of the relay protection zones will be adapted simultaneously as the conditions of the dispersed generation units' change, in order to prevent from the faulty operating of the relay.

Specialized libraries have a significant role in the strategic position of the organization, need to meet the human resource information needs and provide tailored plans for the growth of the organization. Therefore, the purpose of this study is to examine the thematic structure of articles related to specialized libraries during the years 1999-2017 by the method of coword analysis. The present study, with a scientometric approach, has studied the co-occurrence thematic of articles in specialized libraries. A total of 1976 articles (1999-2017) was extracted from his web of science by searching for "Special librar * along with other searches related to specialized libraries." Data analysis after integration using software Netdraw, Ucinet, Ravar PreMap. The findings showed that according to indicators of closeness and betweenness centrality, concepts such as health care, information services, management of specialized libraries, competence, etc. are among the most used topics in this field and most of the scientific outputs related to the United States and England belongs. The results showed that the main focus of the articles on specialized libraries related to human issues in the library. In other words, more attention has been paid to the issues of librarians and library staff. The lack of topics such as resource storage and retrieval, performance measurement, library design, and construction, etc., is a sign of less attention by researchers in these areas.

Load balancing, resource allocation management, and scheduling tasks are the most important challenges in cloud computing. Load balancing is performed at virtual machine level in data centers, while scheduling and allocation of resources are applied on the hosts and virtual machines, respectively. Therefore, the process of allocation of resources to virtual machines in cloud computing and maintaining the load balanced are performed using dragonfly optimization algorithm in this research. The proposed method consists of initializing the algorithm, determining the number of virtual machines and the number of tasks, implementing the dragonfly optimization algorithm, allocating resources, and scheduling the tasks by maintaining the load balanced on virtual machines. Compared to the other methods, improvement of the dragonfly algorithm is approximately 1.34 times more at execution time, 20 milliseconds faster at response time to requests and 6% improvement in term of number of migrations.