Search

Routing protocol for Low Power Lossy Network (RPL) is a proactive network layer routing protocol. It is suitable for Internet of Things (IoT). It is also a source routing protocol. It works by creating a tree based network topology called Directed Acyclic Graph (DAG). Based on the position of the nodes from the root, rank value is allocated for each node. Rank is a unique value given to all the nodes in the DAG. It increases in the downward direction from the root. This value is used for prevention of loops, formation and maintenance of topology. Certain malignant nodes use this rank value and introduce a false routing by forcing the neighbors to route the packets through them. The consequences of this attack are more destructive in nature. It decreases throughput and increases delay in communication. Since the devices in IoT are resources constrained, there is an urge to address rank attack immediately. In this paper, a thorough review on the effects of rank attack and its countermeasures are presented.

Internet of things is a group of objects or devices that establish communication through Internet. Since, the communication is through Internet, more processing power is required in the devices. The IoT devices are heterogeneous and energy constrained. Owing to constrained processing power, the traditional security procedures are not applicable to be executed in the devices. This generates different attacks like user to root, DoS, remote to local and probe attacks. In the existing systems, the classification is carried out to classify different intrusions by using various machine learning algorithms based on KDD dataset. The results were analyzed with various parameters. ML algorithms led to false classifications with medium accuracy. So, it was unable to classify various attacks perfectly. In this paper, the two variants of recurrent neural network such as long short-term memory and bidirectional LSTM are used for the classification of different IoT attacks. The proposed algorithms reduced the false classification with good accuracy. The parameters are computed for the proposed algorithm and are compared with existing machine learning algorithm.

Cryptographic protocols are the backbone of information security. Unfortunately the security of several important components of these protocols can be neglected. This causes violation of personal privacy and threats to democracy. Integration of biometrics with cryptography can overcome this problem. In this paper an enhanced session key agreement protocol which uses the data derived from iris signature is suggested to improve the security of biometric based applications like e-Passport, e-Driving license, etc. The authenticity and security properties of the proposed protocol are analyzed using ProVerif tool and demonstrate it satisfies the intended properties.

Surveys are mainly conducted to obtain valuable information on some criteria from a specified population. But, the survey results often become biased due to non-response of the subjects under study for highly significant attributes. Such non-ignorable missingness need to be treated and the actual values should be retrieved. Many methods have already been proposed for handling missing values in either discrete or continuous attributes. But, there exists a large gap in handling non-ignorable missing values in datasets with mixed attributes. With the intent of addressing this gap, this paper proposes a methodology called as Bayesian Genetic Algorithm (BAGEL) with hybridized Bayesian and Genetic Algorithm principles. In BAGEL, the initial population is generated using Bayesian model and fitness values of the chromosomes are evaluated using Bayesian principles. BAGEL is implemented in real datasets for imputing both discrete and continuous missing values and the imputation accuracy is observed. The experimental results show the superior performance of BAGEL than other standard imputation techniques. Statistical tests conducted to validate the experimental results also prove that BAGEL outperforms at all missing rates from 5% to 50%.

Multilayer Feedforward Neural Network (MFNN) has been administered widely for solving a wide range of supervised pattern recognition tasks. The major problem in the MFNN training phase is its long training time especially when it is trained on very huge training datasets. In this accordance, an enhanced training algorithm called Constant Adaptive Skipping Training (CAST) Algorithm is proposed in this research paper which intensifies on reducing the training time of the MFNN through stochastic manifestation of training datasets. The stochastic manifestation is accomplished by partitioning the training dataset into two completely separate classes, classified and misclassified class, based on the comparison result of the calculated error measure with the threshold value. Only the input samples in the misclassified class are exhibited to the MFNN for training in the next epoch, whereas the correctly classified class is skipped constantly which dynamically reducing the number of training input samples exhibited at every single epoch. Thus decreasing the size of the training dataset constantly can reduce the total training time, thereby speeding up the training process. This CAST algorithm can be merged with any training algorithms used for supervised task, can be used to train the dataset with any number of patterns and also it is very simple to implement. The evaluation of the proposed CAST algorithm is demonstrated effectively using the benchmark datasets Iris, Waveform, Heart Disease and Breast Cancer for different learning rate. Simulation study proved that CAST training algorithm results in faster training than LAST and standard BPN algorithm.

In this Letter, a combined algorithm known as receiver-assisted slow start (RASS) and feedback-assisted recovery (FAR) proposed to address the joint optimisation problems of transmission control protocol (TCP) such as non-optimal initial window (IW) size, slower window growth and spurious window deflation under multi-hop wireless (MHW) networks. The RASS fix the sender's optimum IW size from the receiver's advertised value and replaces the conventional exponential window growth with enhanced window increment mechanism that significantly accelerates the transmission rate. The FAR uses relay router's feedback information to regulate false transmission rate reduction during packet drops. The simulation result reveals that the proposed algorithms had a substantial improvement in throughput and reduction in packet latency over the existing standard New Jersey and non-congestion events algorithms under MHW environment.

Throughput and fairness performance of transmission control protocol (TCP) worsens critically under vehicular environment due to its weakly bounded congestion threshold parameter, rate reduction approach and congestion recovery algorithm. In this paper, a new congestion control algorithm for vehicular networks, called VC-TCP (Vehicular customized -TCP) is proposed to overcome the limitations of the existing independent congestion control approaches. The VC-TCP implements a newer congestion threshold parameter, which regulates the spurious rate reduction due to an abrupt increase in RTT spike value. During the congestion control process, VC-TCP uses a new recovery algorithm and bandwidth utilization based rate reduction, which substantially improves the flow fairness of TCP traffic flows with different bandwidth utilization levels. The throughput, flow fairness, and packet latency performances of VC-TCP is evaluated against the standard Vegas congestion control approach under simulated vehicular scenarios. The simulation outcomes verify the performance improvement of the proposed VC-TCP approach against the standard TCP variants in the different facets of the vehicular environment.

Time series datasets often suffer from the problem of non-ignorable incompleteness which needs careful attention. Even though researchers introduced many methods to handle the missing values, extra effort is needed in searching suitable method for discrete and continuous attributes. This paper proposes a new mechanism called as Bayesian Genetic Algorithm (BAGEL) which is capable of handling missing values in both continuous and discrete attributes in time series datasets using Bayesian analysis and Genetic Algorithms. In BAGEL, Bayesian principles are used to model the data and Genetic Algorithm is used to search the most accurate value that can be estimated from the available data. The method is applied to impute the missing values at different missing rates and the results produced are found to be more accurate when compared with existing methods.

Developing and maintaining Digital Libraries requires substantial investments that are not simply a matter of technological decisions but also organizational issues. While digital libraries hold plenty of promise both now and for the future, they have been slow in taking off. Some digital libraries have either been completely abandoned or they have been put on hold indefinitely. One of the reasons for this predicament is that developers of digital libraries have approached their implementation the same traditional way of building applications, which is also akin to how structures of physical information organizations are built. Digital Libraries with their universal functionality may be even more flexible and reusable, if designed in a service-oriented manner. Such design should allow decreasing the effort of the creation of new digital libraries and the maintenance and scaling of currently existing large installations. Service-oriented architecture offers a better approach to building digital libraries, including streamlining business components, employing reusable services and connecting existing applications to communicate efficiently. The SOA is still a fairly new concept in DL systems. This chapter investigates the applicability of SOA as a fundamental architecture within the system. Its objective is to design a Service-Oriented Architecture for Digital Library System (DLS) using Web Service technology. SORADLS includes different layers which provide primitive services to the library applications built on top of the DLS. DLS techniques of personalization, alert, and caching build SORADLS as services. This architecture provides a fast, safe, convenient, and efficient service to users connected through the Internet.

In this paper, a new queuing mechanism called proactive agile response (PAR) Queue proposed for the multi-hop wireless node that regulates the queue buildup and stabilizes the traffic flows under heavy traffic condition. The PAR uses multilevel packet marking with no additional computational complexity that notifies and controls congestion at its incipient stage. The PAR queuing algorithm supports dual queue approach, the principal buffer store data packets and the auxiliary buffer for control data packets, thereby ensuring no loss of packets due to buffer overflows. The simulation results were clear and convincing that the proposed PAR outperforms the existing standard queue mechanisms with greater improvement in throughput and reduction delay and dropping probability for different traffic conditions.

Transmission control protocol (TCP) performance under multi-hop wireless network heavily depends on retransmission timeout (RTO) estimation algorithm. The existing Jacobson RTO estimation algorithm implemented in the TCP's retransmission (RT) timer collapses under multi-hop wireless environment due to sudden round trip time (RTT) delay spike, leads to the spurious timeout. This paper proposes an enhanced RTO estimation algorithm known as multi-hop RTO (MRTO) intended for the multi-hop wireless environment. The MRTO estimation algorithm computes spike suppression value of the recently received M RTT samples that withstand against the sudden delay spike impact. The simulation results prove that proposed MRTO yields a substantial estimation error and spurious RTO timeout reduction against the Jacobson's RTO estimate and Cauchy Round Trip Time Predictor (CRTTP) under multi-hop wireless network.

Artificial neural network has been extensively consumed training model for solving pattern recognition tasks. However, training a very huge training data set using complex neural network necessitates excessively high training time. In this correspondence, a new fast Linear Adaptive Skipping Training (LAST) algorithm for training artificial neural network (ANN) is instituted. The core essence of this paper is to ameliorate the training speed of ANN by exhibiting only the input samples that do not categorize perfectly in the previous epoch which dynamically reducing the number of input samples exhibited to the network at every single epoch without affecting the network’s accuracy. Thus decreasing the size of the training set can reduce the training time, thereby ameliorating the training speed. This LAST algorithm also determines how many epochs the particular input sample has to skip depending upon the successful classification of that input sample. This LAST algorithm can be incorporated into any supervised training algorithms. Experimental result shows that the training speed attained by LAST algorithm is preferably higher than that of other conventional training algorithms.

Information and Communication Technology (ICT) is one of the fast growing industries that facilitate many latest services to the users and therefore, the number of users is increasing rapidly. The usage of ICT and its life cycle produce hazardous substances that need to be addressed in efficient and green ways. The adoption of green computing involves many improvements and provide energy-efficiency services for data centers, power management and cloud computing. Cloud computing is a highly scalable and cost-effective infrastructure for running Web applications. However, the growing demand of Cloud infrastructure has drastically increased the energy consumption of data centers, which has become a critical issue. Hence, energy-efficient solutions are required to minimize the impact of Cloud environment. E-learning methodology is an example of Green computing. Thus, it is proposed a Green Cloud Computing Architecture for e-Learning Applications that can lower expenses and reduce energy consumption.

Service-Oriented approach is the most appropriate architectural approach that is widely used for building highly large-scale enterprise systems. SORAPES is a serviceoriented reference architecture designed for enterprise elearning systems. This architecture is designed by re-using web services and learning objects for distributed environment. It is notified that the system generate more traffic during authentication/ authorization process and retrieval of information from the database. The problem here is to minimize the traffic during authentication and retrieval ode learning objects process. This traffic will be minimized by centrally managing caching scheme and secured web service requests. This improves the performance and scalability of SORAPES in distributed environment, and hence it is proposed to design reference architecture using proxy and caching to authentication. The objective of this research paper is to design a reference architecture using proxy and caching called RAPESPAC. The RAPESPAC reduces the load on large-scale e-learning applications in distributed environment by centrally managing and caching the web service requests. The important features of RAPESPAC are load balancing, proxy, caching and policy. This architecture could be deployed with confidence in high transaction volume service environments providing high availability of data and reliable security mechanism. General Terms Service-Oriented Computing, E-learning Systems, Distributed Computing.

Pair programming is one of the important practices of a lightweight development methodology namely eXtreme Programming (XP). It emphasizes the practice of two persons work- ing together at a single computer terminal, to design, code and test computer programs. The effects of pair programming on software development in industrial organizations were studied and it was found that pair programming increases the productivity by 15%. This evidence created an interest in amongst community of computer science educators to apply pair programming in educational settings. We have conducted an experiment with the students of computer science courses to compare the learning efficiency of students when they adopt pair programming with that of students using tradi- tional method to do laboratory exercises of short duration. The learning efficiency was measured by evaluating design documents, completion time, and marks obtained in a written test that was con- ducted after every exercise. Our research confirms that the adoption of pair programming improves the design ability, reduces time taken to do a laboratory exercise and increases the knowledge and programming skill.

Cloud computing is an emerging computing model which has evolved as a result of the maturity of underlying prerequisite technologies. There are differences in perspective as to when a set of underlying technologies becomes a “cloud” model. In order to categorize cloud computing services, and to expect some level of consistent characteristics to be associated with the services, cloud adopters need a consistent frame of reference. The Cloud Computing Reference Architecture (CCRA) defines a standard reference architecture and consistent frame of reference for comparing cloud services from different service providers when selecting and deploying cloud services to support their mission requirements. Cloud computing offers information retrieval systems, particularly digital libraries and search engines, a wide variety of options for growth and reduction of maintenance needs and encourages efficient resource use. These features are particularly attractive for digital libraries, repositories, and search engines. The dynamic and elastic provisioning features of a cloud infrastructure allow rapid growth in collection size and support a larger user base, while reducing management issues. Hence, the objective of this chapter is to investigate and design reference architecture to Digital Library Systems using cloud computing with scalability in mind. The proposed reference architecture is called as CORADLS. This architecture accelerates the rate at which library users can get easy, efficient, faster and reliable services in the digital environment. Here, the end user does not have to worry about the resource or disk space in cloud computing.

The Optimized Link State Routing is a proactive link state routing protocol for Mobile Ad hoc Network. The Multipoint Relays nodes which are one-hop neighbors are used as a means for flooding of control and traffic messages into the network. The RFC 3626 does not consider the security issues. The secured OLSR framework consists of security module which provides the security with minimum overhead. The overhead is reduced by minimizing the number of Multipoint Relay nodes in the network and providing security for those selected nodes. The security is provided with the help of threshold cryptography technique. The use of threshold cryptography is to provide security measures by distributing the secret key shares and performing encryption with those shares of secrets. The share can be reconstructed using Lagrange interpolation. The destination could recover only if all or threshold numbers of the secret shares are not compromised. The share update mechanism is also aimed which provide security, when the threshold number of shares are compromised. The main aim is to provide the truthfulness of routing messages, especially Topology Control messages and hence prevent forwarding nodes from varying routing messages while forwarding them. The simulation of the above ideas also confirms the results and provides better security and performance improvement.

In this paper, the authors describe a new architecture for the language faculty of an agent that fulfills the interaction requirements of task delegation. The architecture of the language faculty is based on a conceptualization of the language faculty of an agent and a definition of its internal state paradigm. The new architecture is behavior-management based and possesses self-management properties. This architecture is compared with existing abstract self-management architectures, which examines how the new architecture solves unresolved issues of older models. The architecture description is followed by a case study - Multilingual Natural Language Agent Interface for Mail Service, which illustrates its application.

Language ability is an inevitable aspect of delegation in agents to provide for collaborative natural language interaction (CNLI) during task delegation. To cater to users of different languages, the CNLI is required to be multilingual. When multilingual CNLI has to be provided in a manner characteristic of agents, it leads to realizing two types of autonomies viz. behavior autonomy - to provide CNLI in every language and management autonomy for managing the multiple language competencies. Thereby, the language ability of an agent evolves into a language faculty which encompasses behavior and behavior management autonomies. The existing paradigms for the internal state of agents are only behaviororiented and cannot accommodate the behavior and behavior management autonomies. In this paper, a new paradigm for the internal state of the language faculty of agents consisting of the belief, task and behavior (BTB) abstractions has been proposed, defined and illustrated.

Traditional biometric identification systems require larger memory space for storing unimodal biometric information in the direct form, which suffers leakage of information during transmission of data. Hence there is a necessity to develop an enhanced technique which offers improved protection of biometric data with reduced storage space. Recently, the digital watermarking technique has emerged to be a better solution in addressing these issues. In the proposed paper, an enhanced digital watermarking algorithm which conceals a fingerprint and iris code biometrics in the cover image (face) using Discrete Wavelet Transform-Least Significant Byte technique, which avoids leakage of information and tolerates various attacks on the cover image, is presented. Simulation analysis of the proposed system which stores the biometric information in the indirect form provides complete protection with reduced memory space, and also offers improved authentication accuracy than the existing traditional systems.

The distributed and dynamic nature of grids present a number of challenging problems distinguished from conventional distributed computing by its focus on large-scale resource sharing and high-performance computing. Grid technology facilitates data sharing across many organizations in different geographical locations. Grid enabled applications need to manage and process large amount of data distributed across multiple grid nodes and stored in heterogeneous databases. Data replication is an excellent technique to move and cache data close to users. Replication reduces access latency and bandwidth consumption. It also facilitates load balancing and improves reliability by creating multiple data copies. Selecting a site for the placement of replica is attributed to various factors such as number of requests for a particular file, bandwidth, read/write statistics and location of the resources. In this paper, three replication strategies for dynamic data grids namely best replica site, cost effective site and topology based replication have been proposed and compared with no replication and existing methodologies.

Missing values in databases are more common and if untreated distort the estimates. Numerous methods were developed by researchers to replace the missing values in continuous attributes. The simple methods used are less efficient and the efficient methods are very complex to implement. Hence, to maintain a balance between simplicity and efficiency a new method called Bayesian genetic algorithm BGA is proposed based on genetic algorithm and Bayes theorem for both missing at random MAR and missing completely at random MCAR assumption. Accuracy of BGA is compared with that of mean, kNN and multiple imputation in finding the missing values and the results are studied. BGA produces more accurate results than other methods in four datasets studied at different rates of missingness ranging from 5% to 60%. BGA works better even in large datasets resulting in less biased estimates.

Missing data problem degrades the statistical power of any analysis made in clinical studies. To infer valid results from such studies, suitable method is required to replace the missing values. There is no method which can be universally applicable for handling missing values and the main objective of this paper is to introduce a common method applicable in all cases of missing data. In this paper, Bayesian Genetic Algorithm (BGA) is proposed to effectively impute both missing continuous and discrete values using heuristic search algorithm called genetic algorithm and Bayesian rule. BGA is applied to impute missing values in a real cancer dataset under Missing At Random (MAR) and Missing Completely At Random (MCAR) conditions. For both discrete and continuous attributes, the results show better classification accuracy and RMSE% than many existing methods.