Search

Energy consumption is a significant factor in wireless networks with battery-powered nodes and transceivers as a dominant energy consumer .Also mutual exclusion is a fundamental problem in distributed systems.“Mutual exclusion” allows only one process at a time to access the shared resource and the others have to be locked and wait for their turns.. Retaining mutual exclusion, is proposed as a solution to guarantee the correctness of read/written information from/in the shared resource or critical sections (CS).Wireless Network is an unstable network formed dynamically by a connection of wireless mobile nodes without the use of an existing network substructure. Since energy consumption during communication is a major depletion parameter the number of communication must be reduced as much as possible to achieve extended battery life. Since battery technology does not grow as rapidly as CPU or memory does so there is a strong need for the presence of protocols which are as energy efficient as effective.In this paper we present a new fully distributed token–based mutual-exclusion algorithm for clustered Mobile ad-hoc network with message count reduction purpose which eventuates to power optimization. Therefore we use Raymond's algorithm (which is well-known in stepping down message complexity) within the clusters and perpetual Mobility of token (that is so efficient in high load network such as Manet's) between the clusters.

In the present paper, application of data fusion coupled with three electromagnetic nondestructive technologies (magnetic hysteresis loop, Barkhausen noise, and eddy current) has been explored to achieve higher accuracy and reliability of measurements. The ability of the proposed methodology has been evaluated for determining microstructural changes of D2 cold work tool steel during tempering with focus on decomposition of retained austenite and determination of this phase, quantitatively. The proposed data fusion method takes advantage of ordered weighted averaging aggregation operator and fuzzy defined interval to fuse the data produced by different experiments to obtain more accurate and reliable results. The fusion process also uses redundant information produced by applying different mapping functions modeled by neural networks. The results of the experiments revealed that applying data fusion to the outputs obtained from different experiments produces more accurate results compared to the situation that only outputs of one experiment are used.

Spam can have fake links and trick users into redirecting to fake pages and stealing users' information on these pages. Machine learning methods are typically used to detect spam to distinguish the spam pattern from normal text. A good way to detect spam is to use machine learning techniques that can detect the pattern of spam. While large volumes of text or email need to be considered, these methods take a long time to make using these methods impractical. A good way to accurately and quickly detect large volumes of spam is to use big data processing platforms such as Apache Spark. In this research, an attempt has been made to use feature selection by meta-heuristic methods to increase the speed and efficiency in spam detection. In this method, the data is entered into Apache Spark for final processing to be detected by spam with high-speed distributed learning. Experiments show that using the feature selection mechanism can increase learning accuracy by 98.66%. The processing speed in Apache Spark is about 5.453 times higher than non-Spark systems.

Data mining methods can be used to classify people as sick or healthy and to diagnose diseases. One of the important challenges in diagnosing the disease using learning methods such as multilayer artificial neural network is the optimal selection of its parameters that can be solved to some extent with meta-heuristic methods, but these techniques are not without problems because in these methods meta-heuristic algorithm To a certain extent, it can only reduce the error, and to reduce the machine learning error, in addition to the optimal selection of learning parameters, it is necessary to consider and select the appropriate features. The proposed method has two layers and in the first layer the selection of weight and bias of the multilayer neural network is done using the fruit fly algorithm and in the input layer, the feature selection phase is done using the binary version of the fruit fly optimization algorithm to learn to Do it only on the important features and reduce the problem space and increase the learning speed and accuracy. The results of tests on several disease datasets in the UCI database show that the accuracy, sensitivity, and diagnosis of the proposed method are 98.36%, 98.12%, and 98.08%, respectively. The proposed method is more accurate in diagnosing diabetes than the PSO, FA, SHO, and HHO algorithms.

Recently, the world is experiencing generating the huge amount of data in different domains. Data mining and Data analytics and are the practices used for analyzing data and extracting hidden knowledge. One of the major data mining methods which is used to analysis of data is data clustering. Data clustering ease the extract information from each cluster separately. There are many algorithms used to perform the clustering. One of the most famous algorithms which is used for clustering for more than half a century is k-means. By many optimization and enhancement, K-means still considers as the most popular clustering algorithm which is still being used in various domains. This research attempts to conduct a Systematic literature Review (SLR) to collect, classify, and analyze the primary studies about the different version of k-means clustering algorithm. This SLR gives a means of finding, appraising, and interpreting existing researches pertinent to the topic. By narrowing down the crucial sections of debate, we are hoping to establish a foundation for upcoming researches.

Big Data encompasses huge amounts of raw material which influence multitude of research fields as well as different industries performance such as business, marketing, social network analysis, educational systems, healthcare, IoT, meteorology, fraud detection. It aimed to uncover hidden trends and has prompted a development from a model-driven perspective to a data-driven approach. Among numerous properties of Big Data, datasets of Big Data are identified primary as 3Vs attributes which have high variety, velocity and volume. These provide an invaluable insight and assist in making precise decisions. Analyzing this information and outlining the outcome into helpful data is the method for extricating an incentive from these enormous volumes of datasets. Nevertheless, Big Data containing unique features that cannot be handled and processed using the conventional methods. This has presented a significant challenge to the industry. This research paper presents a general outline of the characteristics of Big Data as well as expounds on the present challenges and limitations in this area. It further discusses the future scope in particular the future direction for Big Data research.

The aim of this paper is to introduce a novel methodology for nondestructive determination of microstructural and mechanical properties (due to the various heat treatments), as well as thickness variations (as a result of corrosion effect) of dual phase steels. The characterizations are based on the variations in the electromagnetic properties extracted from magnetic hysteresis loop and eddy current methods which are coupled with a data fusion system. This study was conducted on six groups of samples (with different thicknesses, from 1 mm to 4 mm) subjected to the various intercritical annealing processes to produce different fractions of martensite/ferrite phases and consequently, changes in hardness, yield strength and ultra tensile strength (UTS). This study proposes a novel soft computing technique to increase accuracy of nondestructive measurements and resolving overlapped NDE outputs related to the various samples. The empirical results indicate that applying the proposed data fusion technique on the two electromagnetic NDE data sets nondestructively, causes an increase in the accuracy and reliability of determining material features including ferrite fraction, hardness, yield strength, UTS, as well as thickness variations.

Variation in chemical composition is one the most important parameters that affects the microstructure and mechanical properties of powder metallurgical (PM) steel parts. Therefore, utilization of nondestructive methods to determine the chemical composition and microstructural/mechanical properties of a given PM steel sample could be a subject of importance. In the present study, capability of three electromagnetic nondestructive methods, including hysteresis loop, Hall effect and eddy current methods has been investigated to determine metallurgical properties in powder metallurgical steel parts. In order to study the variations in the chemical composition, three groups of specimens with different contents of carbon (0–0.9 wt%), copper (0–4.5 wt%) and phosphorus (0–0.45 wt%) were prepared. Microstructure and mechanical properties of the samples were evaluated using microscopic observations and tensile/hardness testing, respectively. In the next step, the relationships among the parameters extracted from the electromagnetic methods and weight percent of alloying elements have been established. Finally, to simultaneously predict the accurate weight percent of carbon, copper and phosphorus, as well as mechanical properties, the electromagnetic outputs were fed into a data fusion system. The new proposed data fusion system is an interval-based ordered weighted averaging method that uses some approximation components to first produce different approximations for each of the outputs and then fuse these approximations to increase accuracy. The study shows that multi-electromagnetic methods coupled to a powerful predictive data fusion system could be used to determine the metallurgical characteristics of the PM steel parts with high precision and reliability.

Data fusion methods are widely used in different application areas especially in decision making. In nondestructive detection of hardness which is one of the most important features of cold work tool steels, these methods can effectively be used to increase accuracy of measurements. In this paper, two data fusion methods have been designed to apply on magnetic Barkhausen noise (MBN) method for determining hardness of D2 tool steel nondestructively and compared with other method of decision making. In the proposed fusion methods, different learners are trained to estimate hardness of the objective materials based on the MBN outputs (height, position and width of the peaks). These learners are neural networks with different structures. The decision of the learners for estimating hardness is fused in a fusion process to make final decision about hardness of the material. The experimental results show that data fusion methods applied on MBN outputs have more effectiveness in comparison with other one for assessment of hardness variations of AISI D2 tempered steel.

This article presents a new method to construct multiple classifier system by making diverse base classifiers using weight tuning. In the method presented, base classifiers are multilayer perceptions which creates diverse base classifiers using a three-step procedure. In the first step, base classifiers are trained for acceptable accuracy. In the second step, a weight tuning process tunes their weights such that each one can distinguish one class of input data from the others with highest possible accuracy. An evolutionary method is used to optimize efficiency of each base classifier to distinguish one class of input data in this step. In the third step, a new method combines the results of the base classifiers. As diversity is measured and monitored throughout the entire procedure, it is measured using a confusion matrix. Superiority of the proposed method is discussed using several known classifier fusion methods and known benchmark datasets.

In this paper we propose to combine two clustering approaches, namely fuzzy and possibilistic c-means. While fuzzy c-means algorithm finds suitable clusters for groups of data points, obtained memberships of data, however, encounters a major deficiency caused by misinterpretation of membership values of data points. Therefore, membership values cannot correctly interpret compatibility or degree to which data points belong to clusters. As a result, noisy data will be misinterpreted by incorrect memberships assigned, as sum of memberships of each noisy data to all clusters is constrained to be equal to 1. To overcome this, a possibilistic approach has been proposed which removes this constraint. It has, however, caused another shortcoming as cluster centers converge to an identical point. Therefore, possibilities cannot correctly interpret the degrees of compatibilities. To correct this problem, a number of works have been carried out which all try to change possibilistic objective function proposed by Krishnapuram and James M. Keller. In this work, a hierarchical approach has been proposed based on properties of both fuzzy and possibilistic approaches to overcome this deficiency. Sensitivities of both methods have been studied together with analyzing results obtained by both methods. Superiority of the proposed method as opposed to conventional possibilistic c-means is shown to be conspicuous.

In this paper, a new classifier fusion method is introduced based on a decision template structure as an extension to Dempster Shafer method. It employs multilayer neural networks as base classifiers. The idea relies on the fact that in a multilayer neural network, behavior of each layer can be a guide for modeling decision-making process. The new decision template based method constructs decision template for each layer of the neural networks including all hidden layers such that a complete model of the base classifiers decision making process is built. In the combiner part, a new strategy based on extension to Dempster Shafer method is introduced. Efficiency of this method is compared with some known benchmark datasets.

This paper discusses some of the salient issues involved in implementing the illusion of a shared-memory programming model across a group of distributed memory processors from a cluster through to an entire GRID. This illusion can be provided by a distributed shared memory (DSM) system implemented by using autonomous agents.

The heuristic methods have been widely developed for solution of complicated optimization methods. Recently hybrid methods that are based on combination of different approaches have shown more potential in this regard. Fuzzy simulation and Particle Swarm Optimization algorithm are integrated to design a hybrid intelligent algorithm to solve the np-hard problem such as travelling salesman problem in efficient and faster way of solutions. The results obtained with the proposed method show its potential in achieving both accuracy and speed in small and medium size problems, compared to many advanced methods.

This research presents a two phases heuristic neural network combined algorithmic approach to solve multiple depot routing problem with heterogeneous vehicles. It has been derived from embedding a heuristic based two level clustering algorithm within a multiple depot vehicle routing problem optimization framework. In logistic applications, customers have priority based on some logistic point of view. The priority levels of customers, affect distribution strategy specially in clustering level. In this research we have developed an integrated vehicle routing problem model using heuristic clustering method with Hopfield network. In the first phase of the algorithm, a high level heuristic clustering is performed to cluster customers serviced by a special depot. Next, a low level clustering is done for each depot to find clusters serviced by a single vehicle. Despite other optimization approaches, which solve case studies involving at most 25 nodes optimally, the proposed algorithm overcomes this limitation by a preprocessing stage by applying clustering on nodes. In this approach, a hierarchical hybrid procedure involving one heuristic and one neural network phases was developed.