Browsing by Author "Guddeti, G."

Now showing 1 - 20 of 50

A gene expression based quality of service aware routing protocol for mobile ad hoc networks
(2013) Kubusada, Y.; Mohan, G.; Manjappa, M.; Guddeti, G.
Mobile Ad Hoc Network (MANET) is a collection of infrastructure less multi-hop wireless mobile nodes which communicate together to achieve the global task. Despite lack of centralized control these mobile nodes still coordinate together to deliver the message to the destination node. MANET is gaining its popularity due to its easy deployment and self-organizing ability. In spite of its unique characteristics, mobility of mobile nodes causes frequent link breakups in MANET and thus makes route setup and maintenance a critical and challenging task. As real time and multimedia applications are increasing, there is a need of an efficient Quality of Service (QoS) aware routing protocol for MANET to support such applications. In the present work, the authors proposed an efficient QoS aware routing protocol for MANET based on upcoming Gene Expression Programming. In the proposed work, the information regarding the availability of resources is managed by a resource management module, which assists in selecting the resource rich path. Further, a theoretical proof is given for the proposed model for its correctness. The results are compared with the state of art artificial neural network and support vector regression methods from the performance evaluation point of view and the results are encouraging. Â© 2013 Springer Science+Business Media.
A hybrid bioinspired algorithm for facial emotion recognition using CSO-GA-PSO-SVM
(Institute of Electrical and Electronics Engineers Inc., 2015) Vivek, T.V.; Guddeti, G.
Human-Computer Interaction gets more natural when the machine can detect human emotions faster and accurate. A lot of research is being carried out in the field of affective computing in order to improve the accuracy with speed. Bio-inspired algorithms for feature extraction and classification stages, has improved accuracy and speed further. In this paper, we propose a hybrid algorithm using CSO (Cat Swarm Optimization) with PSO (Particle Swarm Optimization) and GA (Genetic Algorithm) for emotion recognition (ER). This bio inspired algorithm in conjunction with the support vector machine (SVM) will find an optimal feature set from a bigger set. Results from CK+ (Cohn Kanade) [1] dataset demonstrate that our proposed method using CSO-GA-PSOSVM outperforms Emotion Recognition System with CSOSVM by 10.5% in accuracy. This paper also proposes a new E-Learning [2] system to demonstrate its effectiveness and efficiency in real-time scenario. The proposed algorithm is applied over the facial characteristics captured from students in teaching-learning environment. The optimized feature vector obtained is passed to the SVM classifier for classification. Experimental results yield 99% classification accuracy in a person dependent mode with six basic emotions namely Happy, Sad, Anger, Disgust, Surprise and Neutral. Â© 2015 IEEE.
A novel CBCD approach using MPEG-7 motion activity descriptors
(2011) Roopalakshmi, R.; Guddeti, G.
Motion features contribute significant information about a video content. This paper highlights a novel CBCD (Content-Based Copy Detection) approach, by incorporating several motion activity features. First, we extract both temporal and spatial motion features to describe overall activity of a video sequence. Second, we combine these features in a feasible manner, to generate robust video fingerprints. Third, clustering based pruned search is utilized for similarity matching instead of direct searching of video fingerprints. The proposed system is tested on TRECVID-2007 data set and the results demonstrate the effectiveness of the proposed system against several transformations such as random noise, fast forward, pattern insertion, cropping and picture-inside-picture. Â© 2011 IEEE.
A novel energy efficient resource allocation using hybrid approach of genetic DVFS with bin packing
(Institute of Electrical and Electronics Engineers Inc., 2015) Sharma, N.K.; Guddeti, G.
Increased resources utilization from several clients in a smart computing environment poses a key challenge in allocating optimal energy efficient resources at the data center. Allocation of these optimal resources should be carried out in such a manner that we can reduce the energy consumption of the data center and also avoid the service level agreement (SLA) violation. This paper deals with the development of an energy efficient algorithm for optimal resources allocation at the data center using hybrid approach of the Dynamic Voltage Frequency Scaling (DVFS), Genetic algorithm (GA) and Bin Packing techniques. The performance of the proposed hybrid approach is compared with Genetic Algorithm, DVFS with Bin Packing, DVFS without Bin Packing techniques. Experimental results demonstrate that the proposed energy efficient algorithm consumes 22.4% less energy as compared to the DVFS with Bin Packing technique over a specified workload with 0% SLA violation. Â© 2015 IEEE.
A Novel Method for Disease Recognition and Cure Time Prediction Based on Symptoms
(Institute of Electrical and Electronics Engineers Inc., 2015) Shankar, M.; Pahadia, M.; Srivastava, D.; Ashwin, T.S.; Guddeti, G.
Healthcare is a sector where decisions usually have very high-risk and high-cost associated with them. One bad choice can cost a person's life. With diseases like Swine Flu on the rise, which have symptoms quite similar to common cold, it's very difficult for people to differentiate between medical conditions. We propose a novel method for recognition of diseases and prediction of their cure time based on the symptoms. We do this by assigning different coefficients to each symptom of a disease, and filtering the dataset with the severity score assigned to each symptom by the user. The diseases are identified based on a numerical value calculated in the fashion mentioned above. For predicting the cure time of a disease, we use reinforcement learning. Our algorithm takes into account the similarity between the condition of the current user and other users who have suffered from the same disease, and uses the similarity scores as weights in prediction of cure time. We also predict the current medical condition of user relative to people who have suffered from same disease. Â© 2015 IEEE.
A novel spatio-temporal registration framework for video copy localization based on multimodal features
(2013) Roopalakshmi, R.; Guddeti, G.
Fighting movie piracy requires copy detection followed by the accurate frame alignments of master and copy videos, in order to estimate distortion model and capture location in a theater. Existing research on pirate video registration utilizes only visual features for aligning pirate and master videos, while no effort is made to employ acoustic features. Further, most studies in illegal video registration concentrate on the alignment of watermarked videos, while few attempts are made to address the alignment of non-watermarked sequences. We attempt to solve these issues, by proposing a novel spatio-temporal registration framework that utilizes content-based multimodal features for frame alignments. The proposed scheme includes three stages: first, a video sequence is compactly represented using Speeded Up Robust Features (SURF) and audio spectral signatures; second, sliding window based dynamic time warping (DTW) is employed to compute temporal frame alignments; third, robust SURF descriptors are utilized to generate accurate geometric frame alignments. The results of experiments on three different datasets demonstrate the robustness and efficiency of the proposed method against various video transformations. © 2012 Elsevier B.V.
A novel two-step approach for overlapping community detection in social networks
(Springer-Verlag Wien michaela.bolli@springer.at, 2017) Sarswat, A.; Jami, V.; Guddeti, G.
With the rapid increase in popularity of online social networks, community detection in these networks has become a key aspect of research field. Overlapping community detection is an important NP-hard problem of social network analysis. Modularity-based community detection is one of the most widely used approaches for social network analysis. However, modularity-based community detection technique may fail to resolve small-size communities. Hence, we propose a novel two-step approach for overlapping community detection in social networks. In the first step, modularity density-based hybrid meta-heuristics approach is used to find the disjoint communities and the quality of these disjoint communities can be verified using Silhouette coefficient. In the second step, the quality disjoint communities with low computation cost are used to detect overlapping nodes based on Min-Max Ratio of minimum(indegree, outdegree) to the maximum(indegree, outdegree) values of nodes. We tested the proposed algorithm based on 10 standard community quality metrics along with Silhouette score using seven standard datasets. Experimental results demonstrate that the proposed approach outperforms the current state-of-the-art works in terms of quality and scalability. © 2017, Springer-Verlag GmbH Austria.
A parallel fuzzy C means algorithm for brain tumor segmentation on multiple MRI images
(Springer Verlag service@springer.de, 2013) Ravi, A.; Suvarna, A.; D'Souza, A.; Guddeti, G.; Megha
The Fuzzy C Means (FCM) algorithm has been extensively used in medical image segmentation. But for large data sets the convergence of the FCM algorithm is time consuming and also requires considerable amount of memory. In some real time applications, like Content Based Medical Image Retrieval (CBIR) systems, there is a need to segment a large volume of brain MRI images offline. In this paper, we present an efficient method to cluster data points of all the images at once. The gray level histogram is used in the FCM algorithm to minimize the time for segmentation and the space required. A parallel approach is then applied to further reduce the computation time. The proposed method is found to be almost twice as fast as conventional FCM. Â© 2013 Springer.
A parallel segmentation of brain tumor from magnetic resonance images
(2012) Dessai, V.S.; Arakeri, M.P.; Guddeti, G.
Medical image segmentation is nowadays at the core of medical image analysis and supports computer-aided diagnosis, surgical planning, intra-operative guidance or postoperative assessment. Large amounts of research efforts have been made in developing effective brain MR (magnetic resonance) image tumor segmentation methods in the past years. However algorithms proposed so far are time consuming because it involves lot of mathematical computations. Also serial segmentation of multiple MRI slices (usually required for 3D visualization) takes exponential time. This results in need for improvement in performance as far as the time complexity is concerned. This paper proposes a methodology that incorporates the K-means clustering and morphological operation for parallel segmentation of multiple MRI slices corresponding to single patient. Segmentation of multiple MRI slices for tumor extraction plays major role in 3D (Three Dimensional) visualization and serves as an input for the same. The proposed framework follows SIMD (Single Instruction Multiple Data) model and since the segmentation of individual slice is independent of each other and can be performed in parallel and multithreading definitely speeds up the entire process. Also the framework does not involve any kind of inter-process communication thus the time is saved here as well. Â© 2012 IEEE.
A state transition based approach to recognize gestures using multi level color tracking
(2013) Alse, S.; Eligar, S.; Domanal, S.G.; Guddeti, G.
Gesture recognition is one of the most challenging tasks in Human computer interaction and it has wide range of applications. Here we propose a gesture recognition system which does not involve training the machine in order to detect simple gestures. The proposed technique involves multi level color (color inside color) tracking where region of interest (ROI) is found with respect to the outer color and then with respect to the next inner color and so on. Then the technique involves State transition based approach to recognize gestures where the tracked data is broken down into a sequence of transitions which determine a gesture. This technique is used to develop Jarvis[8], an open source project to control Linux systems using gestures and object tracking. Â© 2013 IEEE.
A unique method to uniformly distribute the load in LEACH and simulation
(2013) Saboo, N.; Bhat, P.P.; Panwar, S.; Yadav, R.; Guddeti, G.
A wireless sensor network (WSN) consists of spatially distributed autonomous sensors to monitor physical or environmental conditions, such as temperature, sound, pressure, etc. and to cooperatively pass their data through the network to a main location. Energy efficiency is the important key for WSN. To lower the energy consumption, the network is divided into several clusters in cluster routing algorithm LEACH (Low Energy Adaptive Clustering Hierarchy). Even though clustering provides energy efficiency, but uneven load distribution still occurs in general. In this paper an approach is presented which divides the deployment region into sub-regions and thereby uniformly distributing the load among the sensor nodes. Different shapes are considered for dividing the deployment region into equal sub-regions. The proposed algorithm is known as enhanced LEACH (e-LEACH). Â© 2013 IEEE.
An approach for color edge detection with automatic threshold detection
(2012) Arpitha, M.D.; Arakeri, M.P.; Guddeti, G.
Edge is an important feature for image segmentation and object detection. Edge detection reduces the amount of data needed to process by removing unnecessary features. Edge detection in color images is more challenging than edge detection in gray-level images. This paper proposes a method for edge detection of color images with automatic threshold detection. The proposed algorithm extracts the edge information of color images in RGB color space with fixed threshold value. The algorithm works on three channels individually and the output is fused to produce one edge map. The algorithm uses the Kuwahara filter to smoothen the image, sobel operator is used for detecting the edge. A new automatic threshold detection method based on histogram data is used for estimating the threshold value. The method is applied for large number of images and the result shows that the algorithm produces effective results when compared to some of the existing edge detection methods. Â© 2012 Springer-Verlag.
An efficient cost optimized scheduling for spot instances in heterogeneous cloud environment
(Elsevier B.V., 2018) Domanal, S.; Guddeti, G.
In this paper, we propose a novel efficient and cost optimized scheduling algorithm for a Bag of Tasks (BoT) on Virtual Machines (VMs). Further, in this paper, we use artificial Neural Network to predict the future values of Spot instances and then validate these predicted values with respect to the current (actual) values of Spot instances. On-Demand and Spot are the key instances which are procured by the cloud customers and hence, in this paper, we use these instances for the cost optimization. The key idea of our proposed algorithm is to efficiently utilize the cloud resources (mainly VMs instances, Central Processing Unit (CPU) and Memory) and also to optimize the cost of executing the BoT in the heterogeneous Infrastructure as a Service (IaaS) based cloud environment. Experimental results demonstrate that our proposed scheduling algorithm outperforms state-of-the-art benchmark algorithms (Round Robin, First Come First Serve, Ant Colony Optimization, Genetic Algorithm, etc.) in terms of Quality of Service (QoS) parameters (Reliability, Time and Cost) while executing the BoT in the heterogeneous cloud environment. Since the obtained results are in the form of ordinal, hence we carried out the statistical analysis on both predicted and actual Spot instances using the Spearman's Rho Test. © 2018 Elsevier B.V.
An intelligent content-based image retrieval system for clinical decision support in brain tumor diagnosis
(Springer London, 2013) Arakeri, M.P.; Guddeti, G.
Accurate diagnosis is crucial for successful treatment of the brain tumor. Accordingly in this paper, we propose an intelligent content-based image retrieval (CBIR) system which retrieves similar pathology bearing magnetic resonance (MR) images of the brain from a medical database to assist the radiologist in the diagnosis of the brain tumor. A single feature vector will not perform well for finding similar images in the medical domain as images within the same disease class differ by severity, density and other such factors. To handle this problem, the proposed CBIR system uses a two-step approach to retrieve similar MR images. The first step classifies the query image as benign or malignant using the features that discriminate the classes. The second step then retrieves the most similar images within the predicted class using the features that distinguish the subclasses. In order to provide faster image retrieval, we propose an indexing method called clustering with principal component analysis (PCA) and KD-tree which groups subclass features into clusters using modified K-means clustering and separately reduces the dimensionality of each cluster using PCA. The reduced feature set is then indexed using a KD-tree. The proposed CBIR system is also made robust against misalignment that occurs during MR image acquisition. Experiments were carried out on a database consisting of 820 MR images of the brain tumor. The experimental results demonstrate the effectiveness of the proposed system and show the viability of clinical application. © 2013, Springer-Verlag London.
Analysis of academic research networks to find collaboration partners
(Springer Verlag service@springer.de, 2016) Asiwal, K.; Suresh, B.K.; Guddeti, G.
Social network analysis has been used for decades to find behavioral patterns and relationships that exist between people in a network. Researchers have been collaborating for centuries with the aim of improving the quality of research, to broaden the scope of problems that they tackle, to speed up the output and to disseminate knowledge across authors. Sometimes it becomes difficult to find the right collaboration partner due to various reasons, the major one being the lack of data about individuals working in their chosen domain in geographically separated locations. In this paper, we explain how social network analysis can be used to help researchers in finding suitable collaboration partners with whom they have not worked in the past but can collaborate in the future. Further, we have considered two different analysis techniques â€“ weighted and nonweighted graph and the results are compared based on the relevance of the outcomes. Â© Springer International Publishing Switzerland 2016.
Analysis of free physical memory in server virtualized system
(Institute of Electrical and Electronics Engineers Inc., 2015) Mohan, R.; Guddeti, G.
Degradation of the performance is the part of any long running software systems. This is due to memory leakage, unreleased file descriptors, round off errors and disk and memory fragmentation. It has been found that the memory leakage is the primary cause of any software performance degradation. In order to predict the software performance degradation, the analysis of the resource usage is essential. Here the free physical memory of a server virtualised system is analysed using time series analysis. Â© 2015 IEEE.
Cognitive network layer in MANETs mobility aware routing protocol
(2012) Zakeerhusen, A.G.; Manjappa, M.; Guddeti, G.
It is intended to add cognition to make cognitive network layer in order to design and develop Quality of Service (QoS) aware adaptive routing protocol in Mobile Adhoc Networks (MANETs). QoS-aware routing is challenging as nodes in the network are free to move, the topology will be changing dynamically. Performance of AODV will be less when nodes in the network are highly mobile. In this paper, Mobility Aware Routing Protocol (MARP) model is proposed to extract a core part in MANET that is stable in terms of mobility of the nodes. This core part is a subset of MANET mobile nodes through which transmission will be done. Here selection of paths through this extracted core can ensure more QoS in time. The MARP model not only provides a better way to discover a QoS but it considers an efficient route maintenance scheme by selecting the route which has more stability as source is having knowledge about other available paths. Since MARP is multipath routing protocol, route maintenance is easy and it robust. By simulation MARP show better performance over existing AODV-on demand routing protocol. Â© 2012 Springer-Verlag.
Cross layer service driven adaptive retry limit for IEEE 802.11 mobile ad-hoc networks
(2011) Manjappa, M.; Guddeti, G.
Traffic in future Mobile Ad-Hoc Network (MANET) is expected to carry a mix of real time multimedia, and non real time file transfer etc. Providing Quality of Service (QoS) for these different applications is difficult and the current research on MANET is choosing the Cross Layer Design for providing QoS. The packet loss due to collision is misinterpreted by MANET as route failure and this triggers route maintenance phase causing unnecessary overhead resulting in low throughput. In this paper, we propose a service driven cross layer model in order to increase the throughput by dynamically adjusting the limits of Request to Send (RTS) retransmission for different flows in the network according to the priority. Simulation is done in NS-2 and the proposed method is compared with IEEE 802.11 DCF MAC using two ad-hoc routing protocols namely AODV and DSR. The results show that the prioritized flow achieves higher throughput over un-prioritized flow when compared to IEEE 802.11 MAC. Â© 2011 IEEE.
Design and evaluation of load balanced termite: A novel load aware bio inspired routing protocol for mobile Ad Hoc network
(Kluwer Academic Publishers, 2014) Manjappa, M.; Guddeti, G.
Bio inspired computing based on Swarm Intelligence is successful in dealing with the networking problems such as routing, congestion and load balancing by finding an optimal path to the destination. Most of the existing bio inspired protocols for MANETs focused only on the routing problem. In this paper, a novel heuristic bio inspired routing with load balancing algorithm referred to as Load Balanced Termite (LB-Termite) is proposed for MANETs by exploiting the salient features of social insect, "Termites". The primary objective of the LB-Termite algorithm is to find the stable nodes and thereby giving preferences for these stable nodes during the path setup; thus finding the reliable route to the destination. The secondary objective of the proposed LB-Termite algorithm is to mitigate the stagnation problem by using pheromone heuristic control method. The simulation results of LB-Termite are compared with other state-of-the-art bio inspired routing algorithms (ACO based Simple Ant Routing Algorithm and the Termite algorithm) and non bio inspired (Ad Hoc on Demand Distance Vector Routing Algorithm) routing protocols for its performance evaluation and the results are found to be encouraging. © 2013 Springer Science+Business Media New York.
Distance based termite algorithm for mobile ad-hoc networks
(2012) Manjappa, M.; Praveenkumar, G.H.; Guddeti, G.
Providing Quality of Service (QoS) in Mobile Ad-Hoc Networks (MANET's) is difficult due to dynamic nature of its topology. Today's research trends show that Swarm Intelligence (SI) can be used effectively to provide QoS in MANET and also MANET is not much explored in the area of SI. Motivated by their self organizing behavior and robustness many routing algorithms have been proposed for both wired and wireless networks. SI routing algorithms are driven by mainly two functions, Pheromone update-decay functions and Forwarding functions. In this paper, a new pheromone update and decay function for Termite algorithm is proposed for MANET which reflects the current context of the network that is the distance between the Mobile Nodes at the time of transmitting the packets. Received Signal Strength (Pr) from Physical Layer is used to find the distance and it is made visible at the Network Layer through Cross Layer Model. The proposed model is simulated and the results are compared with the existing methods and the metric used for the comparison are throughout and control packet overhead. The results show that the new distance based pheromone update and decay methods perform better than the other existing methods. Â© 2012 Springer-Verlag.