مجله رایانش نرم و فناوری اطلاعات
سال دوم شماره 1 (بهار 1392)

The concept of cognitive radio such as sensing free frequency band and dynamic spectrum access can achieve efficient use of the frequency band in future communication systems. One of the important issues discussed in cognitive radio network is resource allocation. Given the competitive character of the user, in this paper we model a cognitive radio network as a noncooperative game where secondary users compete intractably with each other in frequencyselective channels, by maximizing their own information rates subject to the total transmit power and peak transmit power in each frequency band. We express the solution to the problem in Nash equilibrium form and investigate the existence and uniqueness of the Nash equilibrium. Also, to achieve Nash equilibrium of the game, we use an asynchronous distributed algorithm called smoothed asynchronous iterative water-filling algorithm. This means that some users, among others, will have more opportunity to update their own allocated power.

Nowadays one of the aspects of learning is using internet and other information technology products. Already some efforts have been accomplished in using Internet services for teaching and learning different science materials، including religious instructions. One of these efforts was concerned with establishment of question answering systems، where specialists could give answers to the users questions. RASEKHOON is a web site which comprises more than eighteen thousand religious questions with their corresponding answers. In practice، users must look for a question which mostly match with what they are concerned about. The aim of this research is to develop a system that takes a question from a user، then it tries to find the most similar questions in the database to the asked question. The searching procedure benefits from an extension of a language model which is based on intelligent information retrieval and natural language processing techniques. It lets users find answers to theirs questions very rapidly.

In this paper a new method for Multi-Objective Optimization (MOO) has been proposed based on Central Force Optimization algorithm. This method has been called ‘Multi- Objective Central Force Optimization’ algorithm (MOCFO). MOCFO utilizes the concept of ‘Pareto Optimality ’to identify the positions of non-dominated vectors and employs a repository to maintain the positions. The performance of the MOCFO has been evaluated and compared with other optimization techniques which utilize other heuristic algorithms (e.g. particle swarm optimization, and genetic algorithm). The simulation results show that the performance of the proposed MOCFO is comparable to, sometimes better than other MOO techniques. To ensure the true performance of the method presented when opposed with Multi-Objective Optimization Problems, we evaluate it on standard test functions. The final results exhibit the robustness and performance of the so-called method which spans new opportunities of research for the researchers.

Nowadays, Dumpster-Shafer theory of evidence is widely used in data fusion. This theory is a powerful tool for modelling, combination and inference of uncertain, imprecise and incomplete information. The fundamental object of D-S theory is a primitive function called Basic Probability Assignment (BPA), which should be determined based on characteristics of the process. How to determine BPA is still a matter of discussion. In this paper, a novel method is proposed for basic probability assignment. In the proposed method, BPA is determined for each point according to its distances from the cluster centers, by means of fuzzy membership functions. The proposed method is used to determine the switch signal of a multiple MPC, and simulation results show that it considerably reduces the undesired switching. Moreover, the advantages of the reposed method when compared to one of the conventional methods would be discussed.

Edge detection in image processing is one of the main techniques used in segmentation, separation and detection of the special parts of the image. The presence of noise and structural anomaly due to the weak local contrast of the medical images are of the reasons that prevent the current operators from accurate detection of the edge in these images. In this paper, the metaheuristic colony algorithm has been used for edge detection in medical images. Rapid convergence to obtain the optimal solution along with the parameters resistant to initialization has increased the efficiency of the algorithm. In different parts of the image, especially the part with pathological damage, the edge is assumed as ant’s food. Receiving 222 medical images composed of 02 retina images taken from diabetic patients, 02 MRI images as well as 02 microscopic images taken from various medical databases and applying system to them in contrast to such known operators as Canny and Sobel, an acceptable level of accuracy 09. 02%, sensitivity 09. 61% and specificity 09% was separated in the target area from the rest of image. The 00. 90% Kappa coefficient indicates the high reliability factor of system in terms of performance. The use of the current combination method for processing of the images has increased the accuracy even in images with high brightness, rendering the F-Measure significant. The accurate extraction of pathological parts from medical images allows the specialist to determine the disease progression stage, and suggest an appropriate treatment in accordance with the disease growth.

In this paper a method is presented for improving fuzzy recognition of Persian Handwritten Digits. First the training samples for each digit are partitioned into several clusters, and then a fuzzy rule for each cluster is created. The number of clusters for each class (infect for each digit) is optimized using Particle Swarm Optimization (PSO) algorithm. Our approach can improve the recognition rate of a fuzzy algorithm that has only one rule for each digit. The proposed method has increased recognition rate on training data and testing data from 68. 21% and 61. 1 % to 16. 55% and 18. 21%, respectively.

The Quadratic Assignment is a NP-Complete problem, and deterministic algorithms can only solve smaller instances of the problem (at most in size of 21). Thus, developing of Meta heuristic approaches for the problem is worth considering. In this paper, two algorithms have been proposed. The first method is a kind of ICA algorithm and the second one is a hybrid approach which is based on the ICA as well as Object Migration Learning Automaton. Generally, in basic ICA, weaker countries (solutions) are assimilated into stronger solutions as time goes by; therefore, the algorithm may be stuck in local optimum solutions. However, the global optimums can be obtained by exploring the search space; indeed, the learning automaton used in the proposed hybrid method has this responsibility. A number of instances of the standard library of QAPLIB have been used to evaluate performance of the algorithms; in fact, most of these samples are real engineering problems. Experimental results indicate that the newly developed hybrid method is capable of generating highly suitable solutions which a large majority of them are optimal or near optimal; consequently, this algorithm is an effective way to solve the QAP problem.

Automatic signal type identifier plays an important role for modulation identification of satellite signals. Most of the proposed methods didn’t have good performance in low level of signal to noise ratios (SNRs). Also they cant identify more digital modulations. This study investigates the design of an accurate system for identification of digital modulations. First, it is introduced an efficient system that includes two main modules: the feature extraction module and the classifier module. First module extracts a suitable combination of the higher order moments up to eighth, higher order cumulants up to eighth. In the classifier module, an efficient supervised classifier, i.e. radial basis function neural network is proposed. The results show this system has good performance and recognize a lot of digital modulations. However the performance of system degrades at very low SNRs. Also selection of the parameters of the classifier and feature selection is made by trial and error method. The tradeoff between them is a difficult problem. Then at the second fold we have proposed a hybrid intelligent system which an optimization module, i.e. bees algorithm (BA) is considered in the previous system. This module optimizes the classifier design by searching for the best value of the parameters and the best subset of features that feed the classifier. Simulation results show that the proposed hybrid intelligent system has very high identification accuracy even at very low SNRs. This high efficiency is achieved with little features, which have been selected using BA.