|
E-B-M-Based Decision
for Forwarding the Data
in Intermittently Connected MANETs
-- D Jyothi Preshiya and C D Suriyakala
A Mobile Ad hoc Network (MANET) functions as a self-organizing infrastructure-less network. In this network, each node moves independently in any direction but depends on one another to forward its messages to target nodes. Most routing protocols for MANETs are developed with the assumption that not less than one path exists between the source and the target node. Furthermore, it is assumed that all nodes are involved in packet forwarding process. But this assumption does not hold good for all real-time situations because of high mobility and less density of the nodes present in the network and short coverage range of every node. Possibilities arise in the network so that a node could not forward the data to any of the nodes and so it necessitates packets to be stored up in the nodes buffer until it encounters an appropriate node. For deciding the apt receiver node, status of the network and contact statistics of the nodes are required in order to boost up the packet delivery ratio and to trim down the delivery delay and the total number of transmissions. We proposed Encounter-Buffer-Misbehaving (E-B-M) method which takes the best forwarding decision based on the past encounter history and behavior of the nodes. © 2014 IUP. All Rights Reserved.
Classification of Skewed Data:
A Comparative Analysis of the
Performance of Select Classifiers
--Banyabaishali Mohanty and Subhendu Kumar Pani
Many real-world data exhibit significant skewness wherein one class instances outnumber the other class instances. For example, skewed or imbalance class distribution is observed in domains like financial fraud detection, network intrusion detection, etc. The most important fact in these kinds of data is that, compared to the majority class, minority class is often of much significance and interest to the user. However, in such cases, skewed class distribution increases the problems associated with classifying the dataset instances. Usually, a classifier is biased towards the majority class and tends to misclassify the minority class instances, resulting in challenges to data mining. In this paper, we use oversampling to study the classification accuracy of some popular classifiers. The results show that the approach provides satisfactory performance on skewed dataset. © 2014 IUP. All Rights Reserved.
Virtual Appliances-Based Framework
for Regulatory Compliances
in Cloud Data Centers
--Jitendra Singh and Vikas Kumar
In the cloud computing environment, resources are provided and managed by the cloud service provider. The users have no direct control over these third-party resources. Due to this third-party control, the users are always concerned about the safety and security of the data. The user’s concerns towards data privacy and security further increases if the data need to comply with regulatory acts such as HIPAA, GLBA, SAS70 and PCI-DSS. Compliance with these regulatory standards is highly challenging due to the diverse geographical locations of cloud data centers. Any leakage of sensitive information may cause huge penalties or even imprisonment to the cloud provider or the user. In order to enforce the regulatory requirement, a Virtual Appliances (VAPs)-based solution is proposed in the present work. The specific VAP can be executed by the service provider for identification, categorization, routing and storage of the information. The four-layered architecture is suggested for the VAPs, which include: (a) identification; (b) classification; (c) routing; and (d) storage layers. The implementation structure is presented using a framework model along with the algorithm. © 2014 IUP. All Rights Reserved.
Cyber Crimes: Evolution,
Detection and Future Challenges
--Raksha Chouhan
Cyber criminals have new tools at their disposal and are becoming more adaptive than ever. In the present scenario, advent of technological revolution has given broader opportunities and scope to Internet users, but at the same time this has led to the global high-tech cyber crime. India is rated in the top 5 countries affected by cyber crimes, which are gaining momentum from simple e-mail type of crime to serious crime like hacking, phishing, vishing, source code theft, cyber staking, Internet time theft, web jacking and cross-site scripting. Internet has acted as an alternate avenue for the criminals to conduct their activities and launch attacks with relative obscurity. Nowadays, cyber criminals are targeting the social and professional networks and threats are directed at the mobile platforms like smartphones and tablets. Traces of illegal digital activities are often buried in large volumes of data, which are hard to inspect with the aim of detecting offences and collecting evidence. In this paper, an attempt has been made towards evolution of cyber crime, its detection and challenges ahead. The paper also sheds light on the present status of cyber crime, steps taken by the government organizations to avoid cyber crime and also upon international accordance and solutions to deal with the same. © 2014 IUP. All Rights Reserved.
Wireless Sensor Networks:
Applications and Impact
--Monika Kohli and Rohit Tiwari
Nowadays, Wireless Sensor Networks (WSNs) are increasingly becoming popular from all over the world. The applicability of WSN is increasing day-by-day in numerous arenas. Sensed data is being collected, stored and shared using WSNs. WSNs are applicable in various fields like home control, building automation, industrial automation, medical applications, sensor and robots, reconfigurable sensor networks, highway monitoring, military applications, civil and environmental engineering applications, wildfire instrumentation, habitat monitoring and nanoscopic sensor. WSN architecture is basically divided into two types, one is layered and the other is clustered architecture where the functionality of sensor nodes to other nodes is described in detail. In this paper, we emphasize on exploring the applications of WSN in different arenas. In wireless communication, WSN has chartered a new course. © 2014 IUP. All Rights Reserved.
|
