Search

The technological advancement leads to an increase in the usage of the Internet with many applications and connected devices. This increased network size causes increased complexity and creating rooms for the attackers to explore and exploit vulnerabilities to carry out various attacks. As a result upsurge of network attacks can be realized in recent years and is diversified, which can be affirmed by the admittance of various organizations. Varieties of intrusion detection systems (IDSs) have been designed and proposed to tackle such issues based on the misuse-based, anomaly based, and sometimes hybrid techniques. The high rate of network data generation and its enormous volume makes it challenging for IDSs to maintain their efficacy and reliability. This paper discusses a comprehensive understanding of IDS types, six benchmark network datasets, high distributed dimensionality reduction techniques, and classification approaches based on machine learning and deep learning for intrusion detection with their importance to ascertain the efficacy and reliability of IDSs. Furthermore, based on the literature review, a general framework for NIDS has been proposed. At last model for network IDS (NIDS) is designed by following the proposed framework. Achieved accuracy and detection rate of the proposed NIDS model on the UNSW-NB15 dataset are 98.11% and 97.81%, respectively, and achieving better performance than other approaches comparatively.

In the high demand of the cloud computing environment, intrusion detection in a cloud network playing a big role in maintaining the faith of the client(s). Due to the increasing complexity of the c...

Cloud computing provides computing platforms and facilitates to optimize utilization of infrastructure resources, reduces deployment time and increases flexibility. The popularity of cloud computing led to development of interconnected cloud computing environments (ICCE) such as hybrid cloud, inter-cloud, multi-cloud, and federated cloud, enabling the possibilities to share resources among individual clouds. However, individual proprietary technologies and access interfaces employed by cloud service providers made it difficult to share resources. Interoperability and portability are two of the major challenges to be addressed to ensure seamless access and sharing of resources and services. Many cloud service providers have similar service offerings but different access patterns. It is difficult and time consuming for a cloud user to select an appropriate cloud service as per the application’s requirement. Cloud user has to gather information from various cloud service providers and analyze them. Cloud broker has been proposed to address the challenge of cloud users to get best out of cloud provider. Cloud broker is an entity which works as an independent third party between cloud users and cloud providers. Cloud broker negotiates with several cloud providers as per user’s requirements and tries to select the best services. Cloud broker coordinates the sharing of resources and provides interoperability and portability with other cloud providers. In this paper, a comprehensive survey of cloud brokering in interconnected cloud computing environments has been provided. The need and importance of cloud broker has been discussed. The existing architectures and frameworks of Cloud Brokering are reviewed. A comprehensive literature survey of various Cloud Brokering techniques is presented. A taxonomy of Cloud Brokering techniques has been presented and analyzed on the basis of their strengths and weaknesses/limitations. The taxonomy includes pricing, multi-criteria, quality of services, optimization and trust techniques. The techniques are analyzed on various performance metrics. Research challenges and open problems are identified from reviewed techniques. A model for cloud broker is proposed to address identified challenges. We hope that our work will enable researchers to launch and dive deep into Cloud Brokering challenges in interconnected cloud computing environments.

During the last decade, there has been profound progress in the field of visual saliency. However, there still exist various major challenges that hinder the detection performance for scenes with complex composition, presence of additive noise, objects of diverse scale and rotations etc. Generally, images with additive noise have low spatial resolution and blurred edges, which affects the learning capability of the network and causes inaccurate detection. In order to address these issues, in this study, the authors propose a fully convolutional neural network which jointly denoise the input maps by learning edges and contrast details, followed by learning of residing salient details via colour spatial maps in an end-to-end fashion. Their framework employs convolutional layers that use gradient and contrast details of images to denoise the areas with high edge density. After denoising, the denoised images are subjected to salient object detection (SOD) using convolutional layers. The effectiveness of the proposed network is evaluated on benchmark datasets. The experimental results demonstrate the significant performance improvement of the proposed method over state-of-the-art detection techniques.

Internet of Things will serve communities across the different domains of life. The resource of embedded devices and objects working under IoT implementation are constrained in wireless networks. Thus, building a scheme to make full use of energy is key issue for such networks. To achieve energy efficiency, an effective Fuzzy-based network data Fusion Light Weight Protocol (FLWP) is proposed in this article. The innovations of FLWP are as follows: 1) the simulated network's data fusion through fuzzy controller and optimize the energy efficiency of smart tech layer of internet of things (Energy IoT); 2) The optimized reactive route is dynamically adjusted based on fuzzy based prediction accurately from the number of routes provided by base protocol. If the selection accuracy is high, the performance enhances the network quality; 3) FLWP takes full advantage of energy to further enhance target tracking performance by properly selecting reactive routes in the network. Authors evaluated the efficiency of FLWP with simulation-based experiments. FLWP scheme improves the energy efficiency.

Hand gestures are considered as an effective means of communication in the field of Human-computer interaction. However, the design of an efficient hand gesture recognition (HGR) system is still a challenging task owing to a plethora of complexities such as cluttered background, illumination changes, and occlusion in a real-world environment. The paper proposes a lightweight CNN based network named CrossFeat: Multi-scale Cross Feature Aggregation network for explicit hand gesture recognition (HGR). CrossFeat employs multi-scale convolutional layers and preserves the spatial features from the hand gesture region. The use of multi-scale filters: 1 × 1, 3 × 3, 5 × 5 and 7 × 7 allow the network to learn granular and coarse edges from the different regions of the hand gestures. These complementary features enhance the learning ability of the network. Moreover, the cross-layer connectivity enables the gradient information to reach the top layers and prevent it from diminishing in the upstream layers. The proposed network is investigated on three benchmark datasets: ASL Finger Spelling, NUS-I and NUS-II. The experimental results and analysis show that the aggregation of multi-scale and cross features enhances the performance of the HGR system compared to the existing networks.

The P2P networks play a crucial role in communication and data transfer. During recent years the importance of P2P network is increasing due to its scalability and an easy cost-efficient model. From the last few years, a P2P network has also been playing a significant role in real-time data transmission. The features offered by P2P network for video on demand/ live video streaming are getting popular day by day. The key factors which affect the performance of the P2P network are Scheduling and Overlay. Scheduling schemes like push-based, pull-based scheduling schemes that can be further classified into FCFS, rarity based and deadline based scheduling schemes are not sufficient to fulfill the demand of P2P network. So in this paper, we have defined two new scheduling schemes which overcome the shortcomings of the existing schemes. These are priority based scheduling schemes which calculate the priority by combining the different factors which directly affect the performance of a network. This approach is a combination of priority based chunk selection and bandwidth aware peer selection mechanism, which improves the delay between the peers. A comparative analysis of the new priority-based scheduling scheme is done with previous approaches. The simulation results are shown in this paper verify the approach and depict that performance of network and quality of video receiver is improved with the use of priority based scheduling scheme. Performance analysis of scheduling schemes is evaluated using parameters like startup delay, end-to-end delay, playback delay, and frame redundancy, and for the quality of video the distortion, frame loss ratio of video parameter is used.

In the domain of computer vision research, human’s capability to distinguish salient objects from background has become an important ingredient. So far, the researchers are not able to propose an approach that balances the tradeoff between the computation time needed by humans and detection accuracy. This paper highlights an improved salient object detection (SOD) accuracy technique which does not demand high computation time. The accuracy of convex hull enabled active contour technique (CHACT) model is assessed with area under the curve, recall, precision, F-measure, mean absolute error, and computation time on six available publicly image datasets. The computation time of CHACT is comparatively less than many existing approaches and also improved accuracy for SOD.

Cloud computing models are deployed on a compute server whose hardware resources are virtualized to enable multiple virtual machines to run on a single physical system. Several types of virtualization such as bare metal and hosted virtualization are available along with virtualization modes such as full, paravirtualized, hardware-assisted and paravirtualized-hardware-assisted virtualization. Virtual machines are inaccessible from each other when the physical server hardware is abstracted in the full virtualization mode. Physical information such as hard disk drives and server memory are made available in a virtualized environment as a virtual hard disk, vCPU and guest operating system state.

Web related illegal activities are increasing beyond expectation in recent years. Association of National Advertisers has reported that businesses lost more than $6.5 billion in 2017 due to fraudulent activities of ad frauds. Website of Github was down on 28 February, 2018 due to Distributed Denial of Services (DDoS) attack. The attack of DDoS was conducted through memcached servers. Analysis on relevant log data (web access log or ad click log) is required to identify such illegal incidents in digital forensics investigation. An evidence building methodology is proposed and implemented to generate required log data. The proof of concept is provided by coding three scripts, two in python and one in JavaScript. Virtual Machine (VM) on Amazon Web Service (AWS) has been utilized to execute one python script for log separation. Second python script is executed on client database server to store fingerprint of each click or web access. Fingerprint is a generated hash value unique for each device accessing a website or clicking an advertisement (ad). The third JavaScript code is to be embedded in client web page(s) or ad. The verification of log data has also been discussed with the help of fingerprints. The verification process allows regenerating the fingerprints of log data stored in AWS data store. The regenerated fingerprints are being matched with fingerprints stored in client’s database server. The proposed methodology can identify malicious intention of the cloud service provider (CSP) or the investigator or attacker. The proposed methodology can be extended to cloud forensics.

Confidentiality of data is required when it travels over the insecure network such as e-mail transfer using Pretty Good Privacy (PGP), image transfer, etc. The proposed symmetric cryptographic algorithm is suitable for such data transfer. A number of other algorithms can be formulated by combining the existing algorithms to incorporate the algorithms in the relevant applications. The modified algorithm for session key generation is a combination of Affine cipher and hill cipher. A systematic study of the symmetric-key algorithm consisting of multiple keys with regular dependency has been carried out in order to show the efficacy. Security analysis of the proposed algorithm reveals the weakness of the algorithm.

Mudras are considered as spiritual gestures in the religious sense and hold a very important place in the cultural and spiritual space in India. Images are the symbolic representations of divinity in religious artwork and their origins are conveyed through the religions and spiritual beliefs. Such gestures also have some specific meaning in the Buddhist religion. It refers to some of the events in the life of Buddha or denotes special characteristics of the Buddha deities. In recent years, automatic identification of these gestures has gained a greater interest from the machine learning community. This would help to identify the various deities that exist in the Buddhist religion, leading to digital preservation of cultural heritage art. This paper provides a framework that recognizes the Buddhist hand gesture or Hasta Mudra. The morphological features are extracted from the gesture employing geometric parameters. The experimental results show that utilising geometric features and using k-Nearest Neighbor (kNN) as a classifier, an approximately 70% recognition rate is achieved.

Resource Description Framework (RDF) is increasingly being used to model data on the web. RDF model was designed to support easy representation and exchange of information on the web. RDF is queried using SPARQL, a standard query language recommended by W3C. The growth in acceptance of RDF format can be attributed to its flexible and reusable nature. The size of RDF data is steadily increasing as many government organizations and companies are using RDF for data representation and exchange. This resulted in the need for developing distributed RDF frameworks that can efficiently manage RDF data on large scale i.e. Big RDF data. These scalable distributed RDF data management systems competent enough to handle Big RDF data can also be termed as Big RDF frameworks. The proliferation of RDF data has made RDF data management a difficult task. In this survey, we provide an extensive literature on Big RDF frameworks from the aspect of storage, partitioning, indexing, query optimization and processing. A taxonomy of the tools and technologies used for storage and retrieval of Big RDF data in these systems has been presented. The comparative evaluation of some Big RDF frameworks based on query performance and our observations from this evaluation are described. The research challenges identified during the study of these systems are elaborated to suggest promising directions for future research.

Big Data Analytics poses challenges like effective and accurate real-time data mining, lack of suitable tools & techniques and in-memory processing problem. Tree-based ensemble methods (machine learning models) are able to perform such kind of large-scale analytical processing in combination with high-performance cluster computing (special kind of distributed computing) using parallel processing. Random Forest (forest of randomized trees, a tree ensemble) algorithm is considered for the performance evaluation, as tree model supports concurrency and all trees are grown simultaneously in it, so it is a suitable parallel approach with good accuracy, noisy & imbalance dataset handling capability and also it never overfit unlike a single tree model for large dataset. However significant notable improvement over the original approach is available, but some limitation still exists regarding performance and streaming dataset such that performance rate decreases on increasing the compute nodes due to a redundant allocation of feature subsets in the hybrid approach of task & data parallelization and inability to handle stream data. So these performance issues are identified and a problem statement is formulated with an objective to achieve the linear scalable speedup and incremental processing capability of random forest algorithm to perform predictive analytics over massive datasets in the cluster environment.

Use of computational cluster for large-scale Big Data processing has attracted attention as a technology trend for its time efficiency. Modern cluster equipped with latest multi, many-core distributed shared architecture, high speed interconnect and file system, ensures high performance using message passing and multi-threading parallel approaches, also handles batch, micro-batch and stream processing of high dimensional massive dataset but running data-intensive Big Data application on compute-centric cluster imposes challenges to its performance because of several runtime overheads. In order to alleviate these bottlenecks and exploit full potential of the cluster a state of the practice, performance-oriented technical analysis covering all relevant aspects is presented in the context of Terascale Big data processing on TeraFLOPS cluster PARAM-Kanchenjunga, with identification of major factors influencing the performance or sources of these overheads related to computation, communication or IPC, memory, I/O contention, scheduling, load imbalance, synchronization, latency and network jitter; by determining their impact. As existing approaches found insufficient, to achieve possible speedup advance methods with a variety of alternatives as RDMA enabled libraries, PFS, MPI-Integrated extensions, loop tiling, hybrid parallelization are provided to consider for optimization purposes. This paper will assist to prepare performance aware design of experiments and performance modeling.

In the recent years, the efficient management of Cloud Data Center (CDC) resources (computation, storage, and network) is a complex task due to over provisioning and overloading. Furthermore, to form an energy-aware data center with negligible impact on the workload performance is also a critical issue because over provisioned and overload system consumes a considerable amount of electric power. The compute-intensive scientific applications generate complex workload structure (workflow) whose successful execution in terms of energy-performance trade-off is still at its beginning. For this, the simulators are used to analyze and understand the behavior of diverse workload in a distributed environment. WorkflowSim is a popular simulator for workflow processing because of it presents scientific workflows in Directed Cyclic Graphs (DAG) format and offers task clustering (horizontal clustering, vertical clustering, and balanced clustering) for load balancing. In this work, task clustering is used for load balancing and we proposed a Power-aware MAX-MIN VM placement algorithm. Dynamic Voltage Frequency Scaling (DVFS) technique with On-Demand governor and Conservative governor are integrated into Power-aware MAX-MIN VM placement algorithm for reducing power consumption. The implementation results show that energy consumption is reduced considerably in balance, On-Demand, and Conservative approaches with respect to Simple approach.

Cloud computing services can be accessed via browsers or client applications on networked devices such as desktop computers, laptops, tablets and smartphones, which are generally referred to as endpoint devices. Data relevant to forensic investigations may be stored on endpoint devices and/or at cloud service providers. When cloud services are accessed from an endpoint device, several files and folders are created on the device; the data can be accessed by a digital forensic investigator using various tools. An investigator may also use an application programming interface made available by a cloud service provider to obtain forensic information from the cloud related to objects, events and file metadata associated with a cloud user. This chapter presents a taxonomy of the forensic tools used to extract data from endpoint devices and from cloud service providers. The tool taxonomy provides investigators with an easily searchable catalog of tools that can meet their technical requirements during cloud forensic investigations.

Event Detection has gained a lot of attention as a potential research area in the last decade due to emergence and huge adoption of online social media platforms. Though the problem is examined extensively by many researchers, it is the first and foremost important to understand events in detail and then apply detection techniques to identify them. In this paper, we will first study the properties of events and classify them based on these properties. Then, we propose a generic framework for event detection as a possible guideline to develop such systems.

The rapid growth in cloud computing encourages the cloud provider to create more reliable, scalable and efficient cloud environment. For providing such a scalable environment, cloud provider develops large-scale Cloud Data Centers (CDC). Such CDCs consume a large amount of electrical energy for operational purpose and cooling purpose to keep the machines operating at a right temperature, but it also emits a large amount of carbon-dioxide. Thus to decrease the energy consumption, Virtual Machines (VMs) are consolidated to fewer numbers of active servers through VM migration techniques, this leads to Service Level Agreement (SLA) violation and performance degradation. Therefore to retain the energy-performance tradeoff, we should perform an optimum number of VM migrations. There are two cases when the VM migration performed: First when a hot spot is found and Second when a cold spot is found. But in both the cases we do not directly migrate the VM instead first we check whether it is required to migrate the VM. After that, the algorithm finds an appropriate destination host on which the VM will be migrated. In this paper, we implement three approaches to select destination host: First-fit, Best-fit, and Worst-fit. Finally compare the result of a number of VM migration performed, energy consumption and SLA violations with existing one.

Network attacks have become more pervasive in the cyber world. There are various attacks such as denial of service, scanning, privilege escalation that is increasing day by day leading towards the requirement of a more robust and adaptable security techniques. Anomaly detection is the main focus of our paper. Support Vector Machine (SVM) is one of the good classification algorithm applied specially for intrusion detection. However, its performance can be significantly improved when it is applied in integration with other classifiers. In this paper, we have performed a comparative analysis of SVM classifier's performance when it is stacked with other classifiers like BayesNet, AdaBoost, Logistic, IBK, J48, RandomForest, JRip, OneR and SimpleCart. Multi-Classifier algorithm have better classification power when compared to a single classifier algorithm specially for detecting low frequency attacks such as guess password, rootkits, spyware etc. Our preliminary analysis over NSL-KDD'99 dataset shows that stacking of SVM and Random Forest provides the best performance with accuracy of around 97.50% which apparently better than SVM (91.81%).

Salient Object Detection SOD has several applications including image and video compression, video summarization, image segmentation and object discovery etc. Several Methods have been suggested in literature for detecting salient object in digital images. Most of these methods aim at detecting salient objects in images which does not contain any artifact such as noise. In this paper, we have evaluated several salient object detection methods in noisy environment on publicly available ASD Dataset. The performance of the methods is evaluated in terms of Precision, Recall and F-measure and Area under the curve AUC. It has been observed that there is no clear winner but the methods proposed by Liu et al. and Harel et al. are better in comparison to other methods.

Cloud computing offers Information and Communication Technologies (ICT) user's new aspects of ease of services through the Internet. Cloud is a gradual evolution of distributed, cluster, grid and utility computing. Cloud environment attracts the myriads of service providers for development and deployment of their application because it provides ubiquitous and on-demand access to shared resources. The majority of cloud applications like scientific applications are represented as workflow. Workflow is related to the automation of a process that represented by task and data/control information is passed between dependent tasks based on defined set of principles, execution of these tasks require different resources with different Quality of Service parameters. Workflow scheduling is most important concerned in a cloud computing environment because it depends on optimization objectives (like makespan, cost, reliability, security, energy consumption and energy cost, load balancing etc.) of scientific applications. In this review paper, we describe various scientific applications, taxonomy and comprehensive survey of existing workflow scheduling algorithms. We also, classify these algorithms based on nature of the algorithm, optimization model used, optimization objectives and tool used.

Recent years, Cloud Computing attracts the service provider for running applications on large data centers due to their highly available hardware, on-demand provision, and pay-as-you-go concepts. It provides the huge amount of computing power by leverages the virtualization. A virtualization technology is a promising approach to consolidating multiple Virtual machines (VM) onto a minimum number of servers. Dynamic VM provisioning, VM consolidation, and switching servers on and off as required, through all these techniques data centers can sustain the required Quality-of-Service (QoS) while accomplishing higher server utilization and energy efficiency. In our proposed work we can handle the inter-relationship between energy consumption, the number of VM migrations, SLA violation, and performance of the application. The proposed approaches handle the over-utilized servers by migrating the most appropriate VM to the suitable destination server. For this, we propose the VM selection and VM placement approaches. For overload detection, we used the exponential smoothing technique. To implement these approaches, we used cloudsim simulator. Our results show that in energy consumption can be reduced up to 42.3% and the number of VM migration is reduced as well as performances are improved in all approaches in different cases w.r.t. MAD_MU.