Gautam A. Shah

Matrix Approach: Better than Applying Miller’s Equivalents

Abstract The matrix approach is extremely simple and systematic compared to the Miller’s equivalent approach. It is a general approach and the conventional loop and node methods are special cases when there are no controlled sources. It explains how a reciprocal network is converted into a non-reciprocal one due to the presence of controlled sources. It not only gives the exact values for forward gain and input admittance but also gives exact values for reverse gain and output admittance without taking any special precaution.

Miller Equivalents and Their Applications

Miller’s theorems are utilized for approximate as well as exact analysis of both passive and active networks in conjunction with other theorems on a single element or different elements in succession. In this paper, all the four possible Miller equivalents are fully exploited for the exact analysis by applying on different elements simultaneously. This has never been attempted before and may be viewed as an alternate approach for analyzing the networks. The four Miller equivalents are derived using the substitution theorem followed by typical illustrative examples.

A New Fast Radix-2 Decimation-in-Frequency Algorithm for Computing the Discrete Hartley Transform

The radix-2 decimation-in-time fast Hartley transform algorithm for computing the Discrete Hartley Transform (DHT) was introduced by Bracewell. A radix-2 decimation-in-frequency algorithm by Meckelburg and Lipka followed. Prado came up with an in-place version of Bracewells decimation-in-time fast Hartley transform algorithm. A set of fast algorithms for both decimation-in-time and decimation-in-frequency was further developed by Sorenson et al. A new fast radix-2 decimation-in-frequency algorithm for computing the DHT that requires less number of multiplications than those presented by Bracewell, Meckelburg and Lipka, Prado and Sorenson et al is proposed. It exploits the characteristics of the DHT matrix, exhibits stage structures with butterflies similar for each stage and introduces multiplying structures in the signal flow diagram. The operation count for the proposed algorithm is determined. It is verified by implementing the program in C.

Position-based method for computing the elements of the discrete Hartley transform matrix

The elements of the discrete Hartley transform matrix can be computed using the direct method in which each element of the matrix is calculated based on its definition. A position-based method is proposed to simplify the computation. In this method, the characteristics of the discrete Hartley transform matrix are identified and made use of to directly assign values to some elements and compute only a few typical distinct magnitude elements using its definition. These elements are then utilized to obtain the remaining elements based on their positions. An algorithm utilizing this position-based method is developed which is faster in computing the elements than the direct method based on definition.

Compact Dual – Band Antenna for 2.4/5.2/5.8-GHz WLAN Applications

A compact dual – band antenna for 2.4/5.2/5.8-GHz wireless local area network (WLAN) applications ispresented. The antenna consists of two elements, a C shapedelement and a E shaped element. Both of these elements result inexcitation of two resonant modes for dual – band operation. TheC shaped element is fed directly with coaxial probe. The Eshaped element is coupled with ground plane using a connectingcylindrical via. A thorough parametric study is performed tooptimize the antenna performance. The designed antennaprovides bandwidths of about 114MHz (2.32GHz – 2.46GHz) over the lower band and 2.58GHz (4.48GHz – 7.46GHz) over theupper band covering the WLAN frequencies. The antennaradiation pattern shows a broadside pattern over entirebandwidth with maximum gain of about 1.5dBi. The size of theantenna is reduced by 50% in terms of area compared to theprevious research.

No-reference image quality assessment using extreme learning machines

In this paper, a No-Reference Image Quality Assessment (NR-IQA) algorithm is implemented with the help of Extreme Learning Machine (ELM) using spatial and spectral features. ELMs are single hidden layer feed-forward neural networks that provides optimum solution in a single iteration, hence ELMs can be used for performing classification and regression at high speeds. Proposed NR-IQA algorithm can quantify the amount of distortion for images caused by JPEG compression, JPEG2000 compression, Additive White Gaussian Noise, Gaussian Blurring effect and Rayleighs Fast Fading effects. The proposed algorithm is evaluated using LIVE IQA database via Spearmans Ranked Ordered Correlation Coefficient (SROCC) and Root Mean Square Error (RMSE). The proposed algorithm outperforms existing NR-IQA methods.

Design of Wideband Planar Printed Quasi-Yagi Antenna Using Defected Ground Structure

Conventional Yagi antenna provides a unidirectional radiation pattern but is not preferred for wideband applications. The proposed quasi-Yagi antenna consists of a driver dipole, two directors and the ground plane as the reflector. Four extended stubs are added to the ground plane to improve the bandwidth. A U-shaped defect is also introduced in the ground plane which enhances the bandwidth further. Simulation results show that the proposed antenna provides a 111.31 % bandwidth that ranges from 3.94 to 13.83 GHz. The maximum gain offered by the antenna is 5.92 dBi.

A New Position-Based Fast Radix-2 Algorithm for Computing the DHT

The radix-2 decimation-in-time fast Hartley transform algorithm for computing the DHT has been introduced by Bracewell. A set of fast algorithms were further developed by Sorenson et al. A new position-based fast radix-2 decimation-in-time algorithm that requires less number of multiplications than that of Sorenson is proposed. It exploits the characteristics of the DHT matrix and introduces multiplying structures in the signal flow-diagram (SFD). It exhibits an SFD with butterflies similar for each stage. The operation count for the proposed algorithm is determined. It is verified by implementing the program in C.

Current Trends in the Mixed-Mode Signal Processing Architectures

This paper reviews mixed-mode signal processing architectures for computation of discrete signal processing transforms. The Resistance-String and Capacitor-Switched architectures utilize a resistance string and capacitor array, respectively, along with switched-capacitor circuits for their implementation. Switched-Current architecture utilizes switched-current circuits for implementing the processing elements. Programmable block matrix transform architecture uses floating gate circuits for computations. All these mixed-mode architectures make use of different technology and circuits to compute the transforms in real time.