Analysis of Access Delay in Ad Hoc Wireless Networks for Multimedia Applications

Mobile ad hoc networks are infrastructure-less wireless networks; all nodes can quickly share information without using any fixed infrastructure like base station or access point. Wireless ad hoc networks are characterized by frequent topology changes, unreliable wireless channel, network congestion, and resource contention. Multimedia applications usually are bandwidth hungry with stringent delay, jitter, and loss requirements. Designing ad hoc networks which support multimedia applications, hence, is considered a hard task. The hidden and exposed terminal problems are the main which consequently reduces the network capacity. Hidden and exposed nodes reduce the performance of the wireless ad hoc networks. Access delay is the major parameter that is to be taken under consideration. Due to hidden and exposed terminal problems, the network suffers from a serious unfairness problem.

Drug Prediction in Healthcare Using Big Data and Machine Learning

The healthcare system is literally losing patients due to improper diagnosis, accidents, and infections in hospitals alone. To address these challenges, the authors are proposing the drug prediction model that will act as informative guide for patients and help them for taking right medicines for the cure of particular disease. In this chapter, the authors are proposing use of Hadoop distributed file system for the storage of medical datasets related to medicinal drugs. MLLib Library of Apache Spark is to be used for initial data analysis for drug suggestions related to symptoms gathered from particular user. The model will analyze the previous history of patients for any side effects of the drug to be recommended. This proposal will consider weather and maps API from Google as well so that the patients can easily locate the nearby stores where the medicines will be available. It is believed that this proposal of research will surely eradicate the issues by prescribing the optimal drug and its availability by giving the location of the retailer of that drug near the customer.

Application and Impact of Social Network in Modern Society

Social network and its corresponding website permits a client to make a profile, set up an authorized account to create a digital representation of themselves, to select other members of the site as contacts, make connections with them, communicate and engage with these users in different social activities, etc. So, social network includes details of persons, group details, their friends list, contact list, business, affiliations, personal data, personal preferences, and historical information. In this age of smart communication and technology, most of the time people are connected with mobile smart telephones in their work culture, home, office, or any other related places. As they are constantly associated with social systems for long time, they get new posts, messages, and current refreshed news readily available in a flash. This is the constructive part of social networking that individuals consistently remain refreshed with most recent news and innovation. This chapter presents an overview of social network design, various issues, and emerging trends that are evolved simultaneously with modern age. It also presents a detail study on application and impact of social network in modern society as well as exhibits an exhaustive review of security measures in social sites.

Similarity-Based Indices or Metrics for Link Prediction

Link prediction in social network has gained momentum with the inception of machine learning. The social networks are evolving into smart dynamic networks possessing various relevant information about the user. The relationship between users can be approximated by evaluation of similarity between the users. Online social network (OSN) refers to the formulation of association (relationship/links) between users known as nodes. Evolution of OSNs such as Facebook, Twitter, Hi-Fi, LinkedIn has provided a momentum to the growth of such social networks, whereby millions of users are joining it. The online social network evolution has motivated scientists and researchers to analyze the data and information of OSN in order to recommend the future friends. Link prediction is a problem instance of such recommendation systems. Link prediction is basically a phenomenon through which potential links between nodes are identified on a network over the period of time. In this chapter, the authors describe the similarity metrics that further would be instrumental in recognition of future links between nodes.

Online Social Network Analysis

Expansion of online social networks is rapid and furious. Millions of users are appending to it and enriching the nature and behavior, and the information generated has various dimensional properties providing new opportunities and perspective for computation of network properties. The structure of social networks is comprised of nodes and edges whereas users are entities represented by node and relationships designated by edges. Processing of online social networks structural features yields fair knowledge which can be used in many of recommendation and prediction systems. This is referred to as social network analysis, and the features exploited usually are local and global both plays significant role in processing and computation. Local features include properties of nodes like degree of the node (in-degree, out-degree) while global feature process the path between nodes in the entire network. The chapter is an effort in the direction of online social network analysis that explores the basic methods that can be process and analyze the network with a suitable approach to yield knowledge.

Intelligent Medical Diagnostic System for Diabetes

In this chapter, the neuro-fuzzy technique has been used for the diagnosis of different types of diabetes. It has been reported in the literature that triangular membership functions have been deployed for Mamdani and Sugeno fuzzy expert systems that have been used for diagnosis of different types of diabetes. The Gaussian membership functions are expected to give better results. In this context, Gaussian membership functions have been attempted in the neuro-fuzzy system for the diagnosis of different types of diabetes in the research work, and improved results have been obtained in terms of different parameters like sensitivity, specificity, accuracy, precision. Further, for the comparative study, the dataset used for neuro-fuzzy expert system developed in this research work has been considered on Mamdani fuzzy expert system as well as Sugeno fuzzy expert system, and it has been confirmed that the result parameters show better values in the proposed model.

Digital Image Classification Techniques

Image classification is a technique to categorize an image in to given classes on the basis of hidden characteristics or features extracted using image processing. With rapidly growing technology, the size of images is growing. Different categories of images may contain different types of hidden information such as x-ray, CT scan, MRI, pathologies images, remote sensing images, satellite images, and natural scene image captured via digital cameras. In this chapter, the authors have surveyed various articles and books and summarized image classification techniques. There are supervised techniques like KNN and SVM, which classify an image into given classes and unsupervised techniques like K-means and ISODATA for classifying image into a group of clusters. For big images, deep learning networks can be employed that are fast and efficient and also compute hidden features automatically.

Delay Tolerant Networks Architecture, Protocols, and Its Application in Vehicular Ad-Hoc Networks

Due to the high mobility of vehicular nodes in VANETs, there are high chances of partitions in the network. In such a situation, the protocols developed for VANETs cannot work well and an alternative network known as DTN (delay tolerant network) is capable enough to deal with VANET characteristics. The network which does not need any immediate data delivery and can wait for time and delivery of data is known as DTN. The concept of hold and forward the message is exploited by DTN. In this chapter, the authors are providing characteristics, architecture, and applications of delay tolerant vehicular ad-hoc networks.

Deep Learning

Deep learning approaches have been found to be suitable for the agricultural field with successful applications to vegetable infection through plant disease. In this chapter, the authors discuss some widely used deep learning architecture and their practical applications. Nowadays, in many typical applications of machine vision, there is a tendency to replace classical techniques with deep learning algorithms. The benefits are valuable; on one hand, it avoids the need of specialized handcrafted features extractors, and on the other hand, results are not damaged. Moreover, they typically get improved.

Position Independent Mobile User Authentication Using Keystroke Dynamics

In this chapter, a novel technique to authenticate a mobile phone user irrespective of his/her typing position is presented. The user is never always in sitting position while using mobile phone. Thus, it becomes very important to check the accuracy of keystroke dynamics technique while taking input in all positions but authenticating the user irrespective of these positions. Three user positions were considered for input – sitting, walking, and relaxed. The input was taken in uncontrolled environment to get realistic results. Hold time, latency, and motion features using accelerometer data were extracted, and the analysis was done using random forest and KNN classifiers. The accelerometer data provides additional features like mean of all X, Y, and Z axis values. The inclusion of these features improved the results drastically and played a very significant role in determining the user typing behavior. An EER of 4.3% was achieved with a best FAR of 0.9% and an FRR of 15.2%.