International Journal Information and Communication Technology Research
Volume:3 Issue: 2, Spring 2011

Modified active mixer architecture is presented in this paper which improves the gain and noise figure of the mixer considerably. The architecture of the mixer allows wide band operation from very low frequencies up to millimeter – wave regime. Also the proposed architecture makes it possible to use older process options for implementing high frequency mixers with acceptable performance. Simulations using 0.18u RF technology files show the efficiency of the mixer. The technology files are based on BSIM 3(V3.2). Accurate transistor and passive component models are obtained by augmenting the technology files with the results of Electromagnetic field solvers.

A millimeter wave H-plane duplexer with E-plane filters has been designed. The channel filters are connected via a T-junction which is compensated by a cylindrical post and analyzed by integral contour method. Moreover, the junction discontinuity effects have been taken into account. The combination of filters and compensated T-junction allows for low-cost manufacturing. The E-plane filters are analyzed separately, by the mode matching method. The scattering parameters of the combination of filters and T-junction are obtained. The diameter and position of the compensating post and the length of the T-junction arms are optimized in order to achieve the desired performance. This procedure saves computational time and memory. The duplexer has been built for 38GHz and the theory was verified by measurements.

In this paper, a fully distributed model for a single pole single throw traveling wave switch is introduced and important parameters of the switch such as insertion loss, isolation, and reflection coefficient are presented based on the lossy transmission line model of switch. The results of fully distributed model are compared with the semi-distributed model’s results and have good agreement with them. By applying the fully distributed model and calculating various switch’s parameters as a function of the switch length and operating frequency, the optimum switch length and operating frequency are obtained versus the parameters of switch, especially the reflection coefficient and isolation in OFF and ON conditions.

Channel characterization is an important step in the design of wireless communication systems. While channel sounding procedures are a useful method in determining channel behavior, they also require expensive and time consuming procedures and equipment. Ray tracing has been an important substitute for measurement in deterministic channel modeling and characterization of the wireless channel. Much has been done to improve the precision and efficiency of this method for lower frequency bands (generally below 5 GHz) over the years. Recently, with the worldwide announcement of a broad unlicensed band in the millimeter wave spectrum around 60 GHz, a great amount of attention has been paid to this frequency band, previously considered un-utilizable ([1, 2]). However some basic dissimilarity between the 60 GHz and UHF bands has brought about the need for modification of the methods previously used. In this paper the focus has been placed on two propagation mechanisms: diffraction and rough surface scattering, and the impact of each on over-all channel response predictions have been investigated.

Automatic classification of text data has been one of important research topics during recent decades. In this research, a new model based on data fusion techniques is introduced which is used for improving text classification effectiveness. This model has two major components, namely feature fusion and decision fusion; therefore, it is called Feature Decision Fusion (FDF) model. In the feature fusion component, two well-known text feature selection algorithms, Chi-Square (X2) and Information Gain (IG) were used; this component applied Ordered Weighted Averaging (OWA) operator in order to make better feature selection. The second component, Decision fusion component, combined two kinds of results using the Majority Voting (MV) algorithm. The results were obtained with feature fusion and without feature fusion. To evaluate the proposed model, K-Nearest Neighbor (KNN), Decision Tree and Perceptron Neural Network algorithms were used for classifying Rueters-21578 dataset documents. Experiments showed that this model can improve effectiveness of text classification in accordance to both Micro-averaged F1 and Macro-averaged F1 measures.

Localizing sensors in a sensor network is one of the severe bottlenecks that must be dealt with, before exploiting these kinds of networks efficiently. While there has been many techniques and methods proposed for the issue, most of them suffer from low accuracy, or impose extra costs to the network. A Support Vector Machine (SVM) based method has already been proposed which uses machine learning techniques to achieve a fairly accurate estimate of the location of the nodes. In this paper, we propose to use probabilistic SVM, which is more powerful than the existing method. Moreover, an innovative post processing step called ARPoFiL will be proposed that provides even more improvement to the accuracy of the location of the sensor nodes. We will show analytically and experimentally that probabilistic SVM integrated with ARPoFiL completely outperforms the existing method, particularly in sparse networks and rough environments with lots of coverage holes.

Credit scoring is becoming one of the main topics in the banking field. Lending decisions are usually represented as a set of classification tasks in consumer credit markets. In this paper, we have applied a recently introduced rule generator classifier called CORER[1] (Colonial cOmpetitive Rule-based classifiER) to improve the accuracy of credit scoring classification task. The proposed classifier works based on Colonial Competitive Algorithm (CCA). In order to approve the CORER capability in the field of credit scoring, Australian credit real dataset from UCI machine learning repository has been used. To evaluate our classifier, we compared our results with other related well-known classification methods, namely C4.5, Artificial Neural Network, SVM, Linear Regression and Naïve Bayes. Our findings indicate superiority of CORER due to better performance in the credit scoring field. The results also lead us to believe that CORER may have accurate outcome in other applications of banking.

Masquerade attack in computer systems refers to the illegitimate user activities while pretending to be legitimate user. Detection of such attacks is done by discovering significant changes in user’s behavior based on his profile. Profile is built by data produced from mouse, keyboard and other devices. In this paper we propose a practical approach for collecting GUI data and deriving useful parameters included both mouse and keyboard events from Windows OS. We model user identification and masquerade detection as a binary classification problem. Profiling and user classification is accomplished by use of Support Vector Machine (SVM) algorithm. Feature vectors are fed to SVM. The output is behavioral pattern which builds the profile. System is trained by normal behavior and detects deviations from profile. According to the results of implementation the proposed approach ensure detection rate up to 94% with few false alarm.