N. Gopalakrishna Kini

A Torus Embedded Hypercube Scalable Interconnection Network for Parallel Architecture

This paper analyzes an embedded architecture of torus network with the hypercube pertinent to parallel architecture. The product generated from torus and hypercube networks show how good interconnection network can be designed for parallel computation. The advantages of hypercube network and torus topology are used for product network known as Torus embedded hypercube network. A complete design analysis, data routing and comparison of this network with basic networks is given using network parameters.

Torus Embedded Hypercube Interconnection Network: A Comparative Study

analysis and comparison of a product network generated from torus and hypercube networks known as torus embedded hypercube scalable interconnection network suitable for parallel computers is presented in this paper. It is shown here that with minor modifications in architecture of the existing mesh embedded hypercube interconnection network how good a torus embedded hypercube interconnection network could be. Also it has been proved with the computational results that the torus embedded hypercube interconnection network is highly scalable and more efficient in terms of communication.

A Comparative Study of Various Meta-Heuristic Algorithms for Ab Initio Protein Structure Prediction on 2D Hydrophobic-Polar Model

Ab initio protein structure prediction (PSP) models tertiary structures of proteins from its sequence. This is one of the most important and challenging problems in bioinformatics. In the last five decades, many algorithmic approaches have been made to solve the PSP problem. However, it remains unsolvable even for proteins of short sequence. In this review, the reported performances of various meta-heuristic algorithms were compared. Two of the algorithmic settings—protein representation and initialization functions were found to have definite positive influence on the running time and quality of structure. The hybrid of local search and genetic algorithm is recognized to be the best based on the performance. This work provides a chronicle brief on evolution of alternate attempts to solve the PSP problem, and subsequently discusses the merits and demerits of various meta-heuristic approaches to solve the PSP problem.

A Secured Steganography Algorithm for Hiding an Image in an Image

Recent developments in computer security have shown that steganography rather than cryptography is the better method of securing data. A commonly used technique in steganography is the least significant bits (LSB) method, which is, however, vulnerable to attacks due to its simplicity. Nonetheless, it would be difficult or impractical to discern encrypted data if a stego key is added. In this chapter, we show how a carrier image is used to hide another image and present a modified approach for embedding a stego key within it. A 24-bit color carrier image is used to hide the secret image and the same is further used to hide the stego key. We compare the peak signal-to-noise ratio (PSNR) and mean squared error (MSE) and conduct a histogram analysis to determine to what level the stego image is blurred in the carrier image.

A Parallel Algorithm to Hide an Image in an Image for Secured Steganography

Recent developments in computer security have shown that compared to cryptography, steganography is a better way of securing messages. With the advantages offered by parallel computing platforms, a large secret image can be efficiently hidden in another image. This parallelism is achieved in steganography using the OpenCL parallel programming technique. The speed-up improvement obtained is very good with reasonably good output signal quality, even when a large amount of data is processed. The aim of this work is to analyze steganography algorithms and to show how the 24-bit color image of a carrier can be used to hide a secret image. We compare the results by calculating peak signal-to-noise ratio (PSNR), mean squared error (MSE), analysis of histogram and speedup achieved when a large amount of data is processed.

A Novel Framework for Ab Initio Coarse Protein Structure Prediction

Hydrophobic-Polar model is a simplified representation of Protein Structure Prediction (PSP) problem. However, even with the HP model, the PSP problem remains NP-complete. This work proposes a systematic and problem specific design for operators of the evolutionary program which hybrids with local search hill climbing, to efficiently explore the search space of PSP and thereby obtain an optimum conformation. The proposed algorithm achieves this by incorporating the following novel features: (i) new initialization method which generates only valid individuals with (rather than random) better fitness values; (ii) use of probability-based selection operators that limit the local convergence; (iii) use of secondary structure based mutation operator that makes the structure more closely to the laboratory determined structure; and (iv) incorporating all the above-mentioned features developed a complete two-tier framework. The developed framework builds the protein conformation on the square and triangular lattice. The test has been performed using benchmark sequences, and a comparative evaluation is done with various state-of-the-art algorithms. Moreover, in addition to hypothetical test sequences, we have tested protein sequences deposited in protein database repository. It has been observed that the proposed framework has shown superior performance regarding accuracy (fitness value) and speed (number of generations needed to attain the final conformation). The concepts used to enhance the performance are generic and can be used with any other population-based search algorithm such as genetic algorithm, ant colony optimization, and immune algorithm.

Protein structure prediction on 2D square HP lattice with revised fitness function

In order to understand the structure and folding of proteins, Hydrophobic-Polar (HP) model on 2D square lattice is one of the most explored models but parity problem of square lattice make it inefficient for biological applications. This work is dedicated to solve parity issues in 2D square lattice model. This work proposes a revised energy function and presents a case study for protein structure prediction (PSP). A novel approach to evaluate the structure modeling and protein folding problem over HP model is presented in this paper. This evaluation approach may enhance the quality and effectiveness of the computational approaches developed to address the PSP at coarse level.

Intelligent Localization Algorithm for Temperature Monitoring using Wireless Sensor Networks

Sensor Networks (WSN) are increasingly used in monitoring applications like underground water level, building monitoring, environment monitoring etc. The success of these networks depends largely on the ability of the algorithms to run consistently and over a longer duration with least energy consumption. Location stamping is a very important factor in all wireless networks irrespective of whether it is device based or sensor based. It is important to know the location stamping in WSN where the sensor node has sensed the data. In this paper, a novel and experimental implementation of localization algorithm for temperature monitoring system is successfully attempted. The realization of the algorithm is through a combination of Arduino and Xbee, a Zigbee based wireless communication hardware module. The localization is carried out in an indoor environment of a seven storied building with a built area of 237.5 sq. feet with a centralized coordinator approach. Fuzzy logic is used for categorizing the Received Signal Strength Indicator (RSSI) from the coordinator. The algorithm helps to track and identify the location and the individual nodes and also the parameter value recorded over a period of time consistently.