مجله محاسبات و سامانه های توزیع شده
سال چهارم شماره 2 (پیاپی 8، پاییز و زمستان 1400)

Vehicular ad hoc networks (VANETs) are wireless communications networks that use low-tech wireless technologies to create unstable networks among road users. These networks are designed to make the connection between vehicle control and road traffic control. Congestion-based collection optimization is used to discovery near-optimal solutions since network clustering has a difficult NP problem. In this paper, the optimization of improved gray wolf with variable weight-based clustering algorithm for VANETs is presented, which mimics the social behaviors and hunting mechanisms of gray wolves to create efficient clusters. The simulation results show that the proposed framework performs better than the previous works compared to the number of group heads with different communication amplitudes, amount of bulges and network size. Minimizes the cost of routing for the entire network connection by efficiently reducing the number of clusters required. Fewer clusters also reduce the need for resources in the vehicle network.

The growing number of information on the web and the addition of different web pages and websites to this space has made users face problems. These problems appear to users when users are trying to obtain information on a particular topic, and finding all the pages that are suggested to them is a difficult and time consuming process. In the current research, a profile is first created based on the behavioral characteristics of users at different sessions that result from web server logs. These include things like the frequency of user page views, the length of time the user has been on different pages, and the date the page was viewed. We then group them using the clustering method, then fuzzy inference system, extract the fuzzy rules according to the interests of the users and their clusters, and after obtaining the users’ movement patterns, they Insertneural network into vector format Other tools such as bio-algorithms can be useful by obtaining optimal parameters in optimizing predictions and increasing accuracy in fuzzy neural network. The evaluation criteria in this study is accuracy.

This study aimed at identifying the most important factors and barriers to sharing scientific information via social media among Iranian faculty members. A review of the literature on sharing scientific information made it possible to identify some of the factors and barriers to sharing such knowledge. The mixed methodology was used in this study combining qualitative and quantitative methods. The data were collected using online questionnaires. In general, three factors affecting information sharing in the academic environment were identified: organizational culture (including favorable environment for debating, kind of valued information, communication, mutual responsibilities, power distribution and recognition and reward), individual characteristics (including mutual trust, common language, individual time management) and organizational structure (including hierarchy, relationship network, knowledge storage, the transmission of knowledge and type of training for the task). Moreover, four information-sharing barriers were identified and classified in the following categories: individual differences, individual perception of sharing costs, lack of time, lack of recognition and reward, the vision of others, preference for explicit knowledge and resource and infrastructure.

A genetic algorithm has been used to search for artificial intelligence and computing discipline. It assists researchers in finding the most optimized solutions to search problems based on the theory of natural selection and evolutionary biology. Genetic algorithms are robust search algorithm which is appropriate for search in large and complex data sets. There are many ways to produce the individuals in GA through using the crossover and mutation techniques. The final vision of any GA is to maximize fitness function. This paper has proposed a new technique for genetic algorithm uniform crossover by optimizing previous methods. Proposed techniques are developed and tested in the MATLAB platform to see and evaluate the gained result with current crossover techniques. The result shows meaningful improvement in terms of reducing the number of iterations and function evaluations.

The speed and accuracy of data classification are essential factors in the online monitoring of infrasound waves. The available techniques for classifying these data are relatively high accuracy but not high-speed online. In this paper, using the combination of sound and spectral features, the speed of runtime and accuracy of classification have been improved. This method uses spectral entropy, a spectral feature, and the Linear Predictive Coding (LPC) sound features. The proposed approach classifies the infrasound waves generated from the three classes of the rocket launcher, volcano, and shuttle re-entry with a precision of 97% and a runtime of 1.0 seconds that are accurate to the existing methods. The obtained results show that this method has better speed and accuracy rather than existing methods.

According to the WHO (World Health Organization) Cardiovascular (Heart Failure) is a fatal disease that cause estimated 17.9 million people death every year. Heart Disease risk increase due to cholesterol, overweight and hypertension in short due to harmful behavior, heart disease risk increase [1]. Further according to the American Heart Association [2] symptoms(complement) are leg swelling, sleep problem, high heart rate and cough. Diagnosis is difficult due to the common symptoms because these symptoms associate with other diseases. The collection of medical data help physician to diagnosis the disease. Machine learning playing an important role in medical field as diagnosis of disease. Machine learning is used where data is large and difficult to extract useful information from data. In this study, we have used machine learning approach with rapid miner tool for heart disease classification. The experiment results show that performance of Logistic Regression is effective with accuracy of 85.22% than other considered machine learning models.

In the development of a software system, different factors are involved. Due to the complexity of the work, the occurrence of human errors is inevitable. These errors cause the software to behave unpredictably and destructively in the real execution environment. For example, in some programming languages, there is an error called buffer overflow. It may cause unauthorized grants or access to the system in the network and security applications and even endanger user’s privacy or can cause different problems in embedded systems. So software testing is required to increase the reliability and security of the developed software. There are several techniques for testing, among which, the search-based software test (SBST) and the automatic test case generation have a special place. The search-based software test is trying to convert the test problem into an optimization problem and achieve a better solution by searching the problem space. In addition to the benefits of a search-based test, there are challenges. This article provides a comprehensive overview of SBST problems. Understanding challenges is important because it helps scholars find solutions to these challenges or prepare new ideas that will ultimately help improve the quality of future tests and can bring greater security and confidence to the users.

Recently, the growth of convolutional neural networks in various scientific fields can be seen dramatically. The use of various software and hardware techniques in advancing this process provides the platform for increasing research and finding different solutions to increase the efficiency and optimization of this method. One of the important techniques in the field of neural networks is the Generative Adversarial Networks (GAN) and its implementation on FPGA accelerators. In this paper, we will provide an overview of the growth process of convolutional neural networks with using GAN technique and its implementation on FPGA accelerator over two years. In this series we intend to follow some of the most primitive projects starting almost 2017 and by the end of 2018, where significant progress can be made. In this work, we review five papers, the first of which is presented in 2017 and the other four in 2018. The method of comparing these articles is characterized by four distinct perspectives: Optimal utilization of accelerator resources, application of specific techniques, analysis of generated data, and finally the speed of execution in FPGA-based systems is a requirement. Finally, we discuss the advantages and disadvantages of these designs to optimize and improve their performance.

With the accelerated development of Internet finance, electronic funds transfer, and the rapid growth of credit card activity, credit cards play a very important role in every area of ​​life today. There are some risks in this regard that are considered serious threats to both issuers and cardholders. The increasing number of fraudulent credit card transactions forged credit cards and fraudulent use of expired credit cards have led to increased losses. Therefore, finding fraud detection techniques accurately and quickly has become an important topic in current investigations. In this study, after normalizing and reducing the dimensionality of the data using the PCA algorithm, we used the modified perceptron neural network and the grasshopper algorithm to classify the data. In this study, we use the grasshopper algorithm to adjust the weights and biases of neural networks. In the end, we were able to achieve 99.20% accuracy.

The two main alignment problems in genetic sequencing mapping are mapping speed and storage space. In recent decades, effective methods are required to speed up the alignment process and reduce storage space due to increasing DNA sequence data. This study is conducted to provide a hash-based search algorithm to reduce data storage space, which is one of the major issues in hash bases, in addition to increased accuracy and speed. Like hash algorithms, the proposed method uses a hash table to store data, but changes the basis of its work to them, especially the Blast family, and uses a method other than Seed-and-Extend. This increases the speed several times. This method uses the compression technique to reduce the storage space so that the data is stored compressed in the hash table and used during the alignment process, there is no need to decrypt the data in the search and mapping steps, and memory saved is preserved until the end of the process. Besides, unlike most hash methods, this method does not include all possible subsequences of the required length in the hash table when creating the table, and the table is created by passing and recording the position of the existing subsequences of the reference sequence. Another feature of the algorithm is the use of the multiprocessor technique to increase the algorithm execution speed so that the two main and time-consuming steps of the alignment process, one is creating a hash table and the other mapping, are done concurrently.

Fuzzy optimal control problems have been applied in many fields such as electrical engineering, medicine, economics and environments. This fact is because that many of systems are not-crisp form in their modeling equations. In this paper, a class of fuzzy optimal control problems containing a nonlinear system and linear objective function is studied and a more accurate procedure is proposed. and a new piecewise linearization approach are used to solve the optimization problem. At first, the fuzzy nonlinear optimal control problem is obtained and converted to a crisp nonlinear problem by applying defined sets of fuzzy system. Then, the problem is converted to an equivalent linear system using piecewise linearization method. Finally, the accuracy and effectiveness of the proposed approach is confirmed by several numerical examples.

The metal-to-metal connection uses a special welding process, the quality control of which is so important that today the visual control method of destructive analysis has almost replaced the automatic non-destructive detection of welding defects. Among non-destructive methods, this particular method is more comprehensive due to the cheapness of radiography compared to other non-destructive methods and the simplicity of radiographic image analysis. However, the diagnosis of human welding defects may be related to errors due to the low quality of radiographic images. Therefore, today’s radiographic images are analyzed by computer and automated image processing methods. In this work, a clustering method has been used to detect welding defects. The process is to first remove the negative effects such as noise in the image through preprocessing, improve the image quality, and then implement the fuzzy C-means (FCM) algorithm to identify welding defects. In this work, the independent variable is the number of clusters in the fuzzy clustering, and it has been shown that the accuracy of detecting defects in welding images increases as the number of clusters increases. Furthermore, based on the obtained results, the average accuracy of the method for small cracks, large cracks, and void defects is 92.01%, 94.67%, and 99.92%, respectively.