Computer Engineering

- A cloud consists of several elements such as clients, cloudlets, datacenter, distributed servers etc. While Cloud computing is a term, which involves virtualization, distributed computing, networking and software and web services. Centrally, the issue is to establish an effective load balancing in cloud. There are two types of load balancing algorithms available one is Static and other is Dynamic. In this paper we have studied three different Static load balancing algorithms i.e. Round Robin, Throttled, and Optimal. To compare each algorithm we have prepared a set-up and implemented each algorithm. Finally, an analysis has been done about the three implemented algorithms to find the best algorithm in terms of overall response time of cloudlet and Datacenter Processing time of cloudlet.

Abstract: Data mining is emerging technique now a day’s .The main aim of data mining is to find meaningful information from pool of dataset. It is also called knowledge Discovery in Database (KDD). Data mining is technique that extracts the important information from the large amount data base. Clustering is significant key technique in data analysis and data mining application. It is used for combining observed object into groups or cluster such that each group or cluster is homogeneous or compact with respect to characteristics. The clustering is an example of unsupervised learning. A good clustering method will produce high superiority clusters with high intra-class similarity and low inter-class similarity. Clustering algorithms can be categorized into partition-based algorithms, hierarchical-based algorithms, density-based algorithms and grid-based algorithms. The aim of this survey is to provide an analysis of different clustering algorithms in data mining.

People affected with Diabetes Mellitus use Continuous Glucose Monitoring (CGM) devices for monitoring of their blood glucose level. The currently available CGM devices show the current glycemic level of the user and produce alarm whenever the glycemic level exceeds the normal range or is in increasing or decreasing trend. Some devices do prediction of the glucose value some time slots ahead and give an alarm of the impending Hypo/ Hyper glycemia situation. But the true scenario is that the success is only 50% due to false alarms or missing alarms. With the understanding that CGM data do have errors when the rate of change is high, in our study, we have tried the prediction of glucose levels ahead in time without intervention through Auto Regressive Integrated Moving Average (ARIMA) Model with its parameters optimized with Differential Evolution. The Method is validated with Simulated data obtained from the web based Diabetes Educator. First half of the data set is used for training and the remaining half is used for testing. Mean Absolute Difference (MAD) between the predicted and actual glucose values is used as the performance metric. The experiments conducted showed promising results with MAD in the range of 6 to 10.3. The prediction accuracy can be improved by increasing the number of iterations and the optimum selection of scaling factor.

With the increase in the number of mobile applications in the day to day life, essential thing is to keep track as to which ones are safe and which ones aren’t. One can’t judge how safe and true each application is based only on the reviews that are mentioned for each application. Hence it is a need to keep track and develop a system to make sure the applications present are genuine or not. The objective is to develop a system in detecting fraud applications before the user downloads by using machine learning. Sentimental analysis is to help in determining the emotional tones behind words which are expressed in online. This method is useful in monitoring social media and helps to get a brief idea of the public’s opinion on certain issues. The user cannot always get correct or true reviews about the product on the internet. We can check for user’s sentimental comments on multiple application. The reviews may be fake or genuine. Analyzing the rating and reviews together involving both user and admins comments, we can determine whether the application is genuine or not. Using machine learning, the machine is able to learn and analyze the sentiments, emotions about reviews and other texts. The manipulation of review is the key aspects of Application ranking fraud. By using Machine Learning, analyzing reviews and comments can help to determine the correct application for both Android and iOS platforms.

This paper discusses in detail different configurations for comparing the performances of different combinations of commonly used techniques for speech recognition and speaker identification. The Feature extraction techniques considered were MFCC, LPC and Autocorrelation. The feature matching techniques considered were VQ, HMM, DTW and ANN.

Now-a-days the lots of accidents happen on highways due to increase in traffic and also due to a rash driving of the drivers. In many situations the family members or the ambulance and police authority is not informed in time. This result in delaying the help reached to the person suffered due to accident. The purpose of this paper is to find the vehicle where it is and locate the vehicle by means of sending a message using a system which is placed inside of vehicle system .Most of the times we may not be able to find accident location because we don’t know where accident will happen. Real Time Vehicle Accident Detection, Ambulance Rescue and Hospital Management System is designed to avoid such situations. This is an intension to implement an innovative solution for this problem.

In this era of information technology, computer has become indispensable in all fields, including Teaching of English as a Second Language. In India, the TESL has passed through various phases and methods involving technologies, though outdated today were popular then. This shows that throughout the ages, there was a constant search for a new method to realize the needs of the ESL learners. Recently, the concept of Second Language Acquisition has gained significance and linguists try various techniques to minimize the difference between First Language Acquisition and Second Language Learning. Once such method which makes the best use of scientific advancements in TESL research is Computer Assisted Language Learning (CALL).

An adequate blood flow supply is necessary for all organs of the body. Analysis of the blood flow finds its importance in the diagnoses of diseases. There are many techniques for analyzing the blood flow. These techniques are not affordable by the poor people because of their high expense. So we have implemented a technique called Zoom-FFT. This technique is simple and affordable to detect the blood clots and other diseases. Here a specific application will be dealt i.e., ultrasonic blood flow analyzer using ZOOM FFT. The implementation must be achieved with a single VLSI chip in order for the system to be both cost-effective and power efficient and thus widely accepted.

Distributed denial-of-service (DDoS) attacks remain a major security problem, the mitigation of which is very hard especially when it comes to highly distributed botnet-based attacks. The early discovery of these attacks, although challenging, is necessary to protect end-users as well as the expensive network infrastructure resources. Here, we address the problem of DDoS attacks and present the theoretical foundation, architecture, and algorithms of FireCol. The core of FireCol is composed of Intrusion Prevention Systems (IPSs) located at the Internet Service Providers (ISPs) level. The IPSs form virtual protection rings around the hosts to defend and collaborate by exchanging selected traffic information. The evaluation of FireCol using extensive simulations and a real dataset is presented, showing FireCol effectiveness and low overhead, as well as its support for incremental deployment in real network.

The main aim of this project is to design an embedded system for implementing an efficient alcohol and face detection system that will be useful to avoid accidents. Drunk driving face detection with car ignition locking and drowsy driving alert by using raspberrypi aims to change that with automated, transparent , noninvasive alcohol safety check in vehicles. Raspberry pi with dc motor to demonstrate as vehicle engine. If driver is drunk, the processor instantly stops the system ignition by stopping the motor. If alcohol sensor is not giving high alcohol intensity signals, the system lets engine run normally.in additional of this project, automatic turning indication by using gps location tracking and predefined path.