Abstract:HealthCare has been a surging need in India. The inception of Smart India, Smart Villages has further emphasized the need of a Smart & Healthy India. However little reflection of such digitally empowered country has been seen in the field of healthcare. This has further increased the importance of the definition of ubiquity of information and computing technology in healthcare that demand constant surveillance and vigilance of healthcare to predict epidemic outbreak and biological attack. The paper’s primary focus is to analyze and determine the spread of diseases and epidemic in cities/villages. And using this analysis to predict where the next outbreak of epidemic will be. This prediction helps the health authorities to take necessary action in terms of assuring that sufficient resources are available to suffice the need and if possible stop the occurrence of such epidemic by taking necessary actions. To achieve this, we use deep neural network as the heart of our prediction. It receives its training from the past experiences of data which we have collected from hospitals and our spread network. Using this training with our dynamic data it makes predictions as well as adaptively learns from the real time data.

Keywords:epidemic prediction, deep learning, recurrent neural network, machine learning, disease spread network

Abstract:Image segmentation is a process of partitioning a digital image into N regions. In today’s world, Brain tumor detection using image segmentation is fundamental but challenging problem in field of computer vision and image processing due to the diverse image content, image noise, non-uniform object texture and other factors. Accurate detection of size and location of brain tumor plays a vital role in the diagnosis of brain tumor. There are many image segmentation methods available for medical image analysis but the Region based and Clustering techniques are efficient, fast and accurate. This paper presents the two efficient image segmentation algorithms i.e. K-means and region growing techniques for brain MRI images and compare the two algorithms and determine the best one.

Keywords:MRI,Image Segmentation,K-Means,Region growing

Abstract:This paper is in reference with our previous paper entitled “Protection of Key in Private Key Cryptography” published by “International Journal of Advanced Research” in Feb 2017 edition... In this paper we have tried to devise an algorithm that would protect the key in transmission, as in the era of digitization, threats to information security instigate the concerned people to take proactive measures for communication between two communicating parties. Since our major focus is on the symmetric key cryptography, also called as private key cryptography, we have studied different techniques of securing a private key or shared key or secret key that are used to shield information in such systems. The different statistical data being published so far and recent phishing scam which plays with the information transfer by taking away the user’s identity highlights the need to make our security measures strong enough to counter such unwanted accesses. After looking into the functioning of widely accepted protocols and thinking of introducing a few changes to enhance their functionality, we here, are going to focus mainly on application of asymmetric key cryptographic algorithm named RSA algorithm. We have already referred to suggestions by means of several research papers being published so far related to our area of interest such as N-Prime RSA, magic rectangle, etc. We have tried to present a method taking one step ahead in the available sketch and without diminishing the basic essence of the algorithm. We also refer to some examples and its application as thought of in its modified form to further support our idea.

Keywords:Cryptography, Key, Private Key, Public key, Asymmetric or Public Key Cryptography, Symmetric or Private Key Cryptography

Abstract:Internet of Things (IoT) is global network of “smart device” that can sense and interrelate with their environment using the internet for their communication and interaction with users and other systems. The traditional method of testing Turbidity, PH & Temperature is to collect samples manually and then send them to laboratory for analysis. However, it has been unable to meet the demands of water quality monitoring today. By manually monitoring the water quality is very time consuming as well as its complicated to get water parameter in real time. To overcome this a set of Monitoring of Turbidity, PH & Temperature of Water quality are going to be placed in the reservoir as the variations in these parameter will get to know the presence of pollutants. In this paper a new approach has been adopted to analyze the quality of the water under IoT environment. The proposed system contains data acquisitions node, controlling unit as ARM, base station and monitoring unit, where all these units are connected to each other .Data can be collected from various sensor and each sensors has its own way to read the data about physical phenomenon ,so necessity to write diverse code for each sensor for data collection procedures. Meanwhile to control and to do some pre-processing work MCU unit has been adopted, after data collection from various sensor such as its Temperature, Turbidity, PH, etc at base station is sent to display on monitoring node and also compared with standard values .By continually monitoring water quality people get awareness locality water condition so we can minimize its critical effect on human life ,filed, animals etc with high frequency.

Keywords:Internet of things (IoT), WSN, MCU, Communication standards, Temperature sensor, level sensor, Turbidity sensor

Abstract:A wormhole assault is especially hurtful against steering in sensor systems where an assailant gets bundles at one area in the system, passages and afterward replays them at another remote area in the system. A wormhole assault can be effortlessly propelled by an aggressor without bargaining any sensor hubs. Since a large portion of the steering conventions don't have systems to shield the system against wormhole assaults, the course demand can be burrowed to the objective region by the aggressor through wormholes. Along these lines, the sensor hubs in the objective zone construct the course through the assailant. Afterward, the assailant can alter the information, messages, or specifically forward information messages to disturb the elements of the sensor arrange. In this work we show another approach to identify wormhole assaults in MANET. Our location is basic and productive. We require neither GPS gadget, nor clock synchronization which are the fundamental restrictions of the other existing arrangements. Also, our recognition can be effortlessly actualized convention or in any neighbor revelation convention for MANET. We don't present any new messages. Along these lines, the overhead of the arrangement is restricted to the additional data (timestamps) attached to the Hello messages.

Keywords:Mobile Ad-hoc Network (MANET), Wormhole attack, Malicious node