Majlesi Journal of Electrical Engineering
Volume:2 Issue: 1, 2008

In this paper a fully integrated CMOS Distributed Amplifier is presented. This DA can be used for broadband optical and wireless communication applications. A four stage cascode DA is designed and optimized. A critical problem in CMOS RFIC design is the parasitic elements of transistors and inductors and this problem makes handed design methodology complex. Here a CAD tool underpins the parasitic-aware methodology used to optimize the design including all on-chip active and passive devices and bias voltages. Optimization is a new method based on Distributed Pareto-based Multi-Objective Genetic Algorithm that is introduced for RFIC design optimization. The optimization system is parasitic-aware and simulation-based. Through a link between HSPICE and MATLAB, all transistor sizes, bias voltages and number of turns and diameter of inductors are proposed by CAD and then circuit, with these values are simulated by Hspice-RF. The output parameters, such as gain, bandwidth, S11, S22 and power are extracted from output file and the area of chip is calculated separately. This output parameters are used as cost functions for creating next generation. This algorithm is implemented by Matlab and simulated by Hspice-RF with TSMC 0.18u CMOS technology.

Due to expansion of photonic crystal fiber (PCF) based optical devices in optical communication networks, the connection process between conventional optical fibers (COF) as transmission medium and these devices has attracted the researchers. Considering the splice joint between the COF and PCF, in this paper, the fundamental core and cladding modes are analyzed by scalar effective index method (SEIM) and vectorial effective index method (VEIM), and the results are compared with finite difference frequency domain (FDFD) method. Then by using Gaussian approximation, the effects of bending and transverse misalignment at splice joint are analyzed. Based on the obtained results, for the first time, an approach to optimize the attenuation of the splice joint between dissimilar optical fibers is presented.

During the last few years, evaluation of MIMO systems capacity has been expanded, due to more bandwidth demands. In this paper, after reviewing the previous works, a new formula for MIMO systems capacity, in the case of uniformly distributed channel error is given. While in previous works, Gaussian distribution is the preliminary assumption of almost all, we have switched to uniform case, because the channel error comes from the quantization techniques used in receiver to send the information of channel through a partial feedback path to the transmitter and these techniques in nature make uniform distribution error.

A complimentary topology is used incorporating TSMC 0.18 mm process to design a CMOS VCO with the center frequency of 5.4 GHz. Simulation results showed tuning range of 13%. The phase noise at 1 MHz offset was measured to be -118.7 dBc/Hz. The VCO core power consumption was 3.3 mW when the power supply voltage was set to 1.5 V. Simulation results verified that the designed structure meets the IEEE802.11a requirements.

Electrical motor with high performance and maintenance free is very atteractive for the special application in the case of limited energy sources and long operating board. Permenant magnet brush DC motor is very good candidate for such specific applications. In this paper,at first a mathematical model of the motor is persented then, a method for speed control of the motor is developed by using vector (field oriented) control method. The persented method is simulated by PWM switching in Matlab software environment. A control unit is fabricated based on the presented method. moreover the result of experiment is compared with the simulation result. The comparison shows that the presented method had a high efficiency in control of BLDC motor.

One of the methods in creating disturbance in radar is to apply appealing jammer. These jammers send strong signals like target-return signals into radar. If these signals enter from main lobe of radar antenna, create false targets in different ranges. Besides if they enter from secondary lobes, radar receiver is not able to distinguish between signals entered from primary and secondary lobes, and they are interpreted as signals of primary lobes, consequently some false targets are detected in different angles. In this paper, with considering the method that is used in [Shnidman, D.A and Toumodge, S., 2002], we simulate an optimum method to block secondary lobe jammers. Lobe disturbance from primary radar signal, done by applying pulse calculations.

We investigate the effect of ASE(Amplified Spontaneous Emission) on the gain modulation, that also referred to overmodulation. The gain modulation is the low-frequency (kHz) amplitude modulation of the EDFA pump and the communication signal used for propagating line monitoring information. We develop the model of Novak and Moesle (2002), by including ASE, that they neglected. The derivation of an analytical model, for EDFA overmodulation response, has been presented. This model provides analytical expressions for the pump and input signal overmodulation responses, respectively. These expressions describe the output signal modulation index amplitude and phase, assuming small sinusoidal steady-state oscillations of the mean pump or input signal power. In this paper we show that ASE has some effects on predictions in the high gain/low saturation regime.

Language identification (LID) in speech signals is an important classification task. In this paper Persian language verification is proposed and evaluated. The system is developed by using Gaussian mixture models as a basic system for tokenizing and a Neural Network as the backend processor. Gaussian Mixture Models can be utilized to model the distribution of feature vector in speech signals for classification. We gathered our language identification corpus from different Satellite TV channels. The results are presented for a system using the GMM Tokenizer in combining with Neural Network. The results of GMM-NN system compared with GMM-Tokenizer system. It is shown that using the Neural Network as the backend processor improves the results significantly.