مجله رایانش نرم و فناوری اطلاعات
سال چهارم شماره 1 (بهار 1394)

Department of Elec. Eng., Faculty of Eng., University of Birjand, Birjand, Iran Abstract–Inclined Planes Optimization (IPO) algorithm is a novel search and optimization technique. In this paper, an unsupervised image clustering technique (called IPO Clustering) is proposed. In this algorithm each agent is called ball and includes the information of the cluster centers, which are assigned randomly in the first iteration of the proposed algorithm. The values of the balls are updated and optimized to reach the stopping criteria. The fitness value of each ball is defined based on the inter-class and intra-class distances. Extensive experimental results show the effectiveness of the proposed clustering algorithm.

An important class of wideband communication signals exhibits cyclostationarity. Using this property, some deficiencies in conventional MUSIC direction-of-arrival (DOA) estimation algorithm including the inability to eliminate the interference, can be removed. Hence, in order to use this property, this paper is focused on wideband cyclic-based MUSIC algorithms and four wellknown algorithms, conventional MUSIC, averaged cyclic MUSIC (ACM), extended wideband cyclic MUSIC (EWCM) and generalized wideband cyclic MUSIC (GWCM) are investigated. Considering wideband desired and interference signals impinging on uniform linear array (ULA) from far field, the performance of different algorithms in estimating the DOA of desired signals and eliminating the interference signals is evaluated. According to this fact that cyclic correlation within a certain range of time delay has large values, appropriate value of averaging window for simulations is obtained 𝜏 = [−01 + 01]. Moreover, the simulation results in MATLAB software show that: 1. By choosing the proper cyclic frequency, interference can be removed. In this research, by evaluating the root mean square error (RMSE) of algorithms, appropriate value of cyclic frequency is obtained 𝛼 = −𝑟𝑏. 5. DOA estimation depends on frequency. Therefore, those methods that take into account the frequency offer higher performance based on spatial spectrum and RMSE. 0. Increasing the desired and interference signals, conventional MUSIC algorithm is unable to eliminate the interference. Also, it does not well estimate DOAs of desired and interference signals.

Recently, improving energy efficiency and increasing network capacity have been attractive in multi-radio multi-channel networks. In these networks, several methods are introduced to improve the network capacity. In this paper, we propose a method considering both interference and energy consumption in communication links. This method consists of two phases: in the first one, an optimized energy consumption sub graph, Gabriel graph, is constructed on the nodes of the network graph. Then, in the second phase, to increase the network capacity, available channels are allocated efficiently. Furthermore, in this paper, the channel conditions are considered on the communication links. With this assumption, the proposed scheme provides a more transmission rate in the links with the better conditions and consequently more probability reuse of these links. Generally speaking, introduced topology constructs a graph with the minimum energy consumption and minimum interference.

Studying the resting state functional magnetic resonance imaging (rs-fMRI) has been an important issue for diagnosing of the diseases related to brain function, nowadays. This paper proposes a novel method for diagnosis of the Attention-deficit/hyperactivity disorder (ADHD) diseases that includes three main modules. The first module is the feature extraction that extracts the correlation between the time series of the signals of the different brain regions. In the second module the SVM is used as the classifier. The third module is an optimization algorithm, i.e. particle swarm optimization algorithm (PSO). PSO chooses the appropriate features by changing the length of its particles. The results show that the proposed method has high classification accuracy in comparison with others.

In this paper an automated and robust method for detection and tracking of moving target with assumption of camera mobility using scale invariant feature transform is proposed. By automated registration technique of current frame on the previous frame, a panorama based background model of scene is generated. Then for real-time panorama-based background subtraction, frame registration into the panorama and global homographic matrix estimation are used. After that for each new frame, moving targets are detected and background model is updated and the key points where detected in the boundary region around the target are used for tracking. By matching description vectors of at least three key points where extracted from scale invariant feature transform of current frame to previous frame, initial estimation of target location on current frame are obtained. In the next step, more accurate estimation of target location on current frame is obtained by using Kalman filter. This approach is able to be used in different statues, such as variable climate condition, closing or faring the target, partial and global occultation. Comparison to expert operator, this approach improved performance results and achieved high accuracy in detection and tracking of aerial and ground target.

It is important to determine the dose of drugs for treatment of cancer with a view to expense and also side effects entered on the patients. In this paper with profiting of mathematical model one of the kinds of blood cancer named chronic myelogenous leukemia and with using of experiences of blood cancer specialist has been designed Fuzzy controller that determine necessary drug dose for treatment of this disease in a period of 521 days. This method is experimented on 0 patients with disease diverse condition and different parameters and the result are shown that this method contrary to present methods has very proper robustness for treatment of different patients, meanwhile the amount of consumption drugs in comparison with other papers has been reduced dramatically.

The university course-timetabling problem is assigning the available time slots, resources and other facilities in an educational environment like a university, such that the overall utilities of the professors, teacher assistants, students and employees are satisfied as high as possible and the rules or constraints of the university are not violated. In this paper, considering the criteria and conditions of the Tabriz branch of Islamic Azad University, a new multi-objective linear mathematical programming model for this problem with four objectives is proposed and it is solved using fuzzy goal programming approach to aggregate the objective functions in order to maximize the minimum utility of the objective functions. Due to the NP-Hard complexity of the proposed mathematical model, a Genetic Algorithm (GA) is also developed for solving the model. Two different departments from the university are selected as case studies and the related manual data are given as the input data to the proposed model. Comparing the obtained results on the real data adopted from departments with the results obtained by head of the department using the manual traditional methods shows the superiority of the proposed model in obtaining the better solutions having more utilities.

Extraction, transformation and loading (ETL) of data sources changes are an important subject in the data warehousing field. During ETL process data are extracted from operational data sources, and after the extraction, these data can be transformed and loaded into the data warehouse. Because operational data sources change over time, it is required that the process of the data warehouse refreshment be done in the maintenance phase and changed data loaded into the data warehouse. Incremental loading is a widely used method to load data in data warehouses from the respective source systems. The main problem of this method is that if the number of changes increases then it will become time consuming. Usually, these processes must be completed in a certain time window; thus, it is necessary to optimize their execution time. The main purpose of a proposed method in this paper is to optimize time of changes loading by reducing number of log records and changing possible delete and update operations into insert operation in incremental loading. According to the results of tests conducted by real and test data and the comparisons, a significant improvement has been achieved due to the proposed data loading method of this paper (53.02 on real data).

Tracking has considered many years ago and it has been concerned as an unsolved problem. This paper presents a methods using Genetic algorithm. The proposed algorithm uses background subtraction and multi scale LBP to detect moving target region. We have used Chi-square Criterion to calculate the similarity between two regions. The simulation results shows that the proposed method is efficient in accuracy and speed. In each frame the average time for calculating the position of the vehicle is 0.33 seconds while the sequential search time is 02.15 seconds.