Search

The strong growth of communication and storage gives rise to the significantly increasing demand for collecting, storing, and sharing a large amount of data on networks. This is further enhanced by the data-driven market, where everyone wants to access other parties' data. Data is the bedrock of today's technologies and researchers, especially in machine learning and deep learning. The business value of organizations is also massively data-dependent. Recent studies and industry applications can apply analytic techniques for exploiting data, or Internet users can exchange data on social networks or peer-to-peer (P2P) networks. However, sharing secured data is a challenging problem that attracts much attention from researchers. Sharing secure data with a group of users using P2P applications faces the unavailability problem of peer nodes. Thus users cannot certify and download the protected data. This affects a P2P-based application class of sharing and storing online services or customer-to-customer e-commerce applications. This article proposes a solution for sharing secured data on P2P-based applications using blockchain and attribute-based encryption. The attribute-based encryption guarantees sharing keys among a group of users, while blockchain guarantees keys distribution. We have simulated the proposed solution on the mobile peer-to-peer network that provides services for sharing and storing data securely.

Processing the 3D objects is a research field of computer graphics that can be applied in 3D simulation, 3D image processing, game industry, etc. This is one of the mathematics-based research fields where geometrical knowledge is the background of existing methods. Geometric modeling is a branch of this research field based on the applied mathematics and computational geometry. It studies methods and solutions for the mathematical description of shapes or realistic objects in computer science or computer aided design. The rapid development of advanced techniques, like 3D scanners, can be used to support obtaining scanned data of any type of the real objects with different methods and solutions. This article investigates the studies based on the background of geometric modeling for processing the 3D Objects. Our contribution is focused on presentation of a completed method that is an integrated approach consisting of our previously proposed methods to reconstruct the 3D object from a 3D point cloud dataset. The method consists of four main steps, as follows: (1) Obtaining data describing the 3D objects in the real world; (2) Processing and simplifying the obtained data; (3) Reconstructing the 3D objects by filling the holes on the surface; and (4) Meshing and visualizing the 3D objects in the application. The comparison and discussion between the methods have shown the advantages of the proposed method and its application in practice.

Automating fault detection in communication networks and distributed systems is a challenging process that usually requires the involvement of supporting tools and the expertise of system operators. Automated event monitoring and correlating systems produce event data that is forwarded to system operators for analyzing error events and creating fault reports. Machine learning methods help not only analyzing event data more precisely but also forecasting possible error events by learning from existing faults. This study introduces an automated fault detection system that assists system operators in detecting and forecasting faults. This system is characterized by the capability of exploiting bug knowledge resources at various online repositories, log events and status parameters from the monitored system; and applying bug analysis and event filtering methods for evaluating events and forecasting faults. The system contains a fault data model to collect bug reports, a feature and semantic filtering method to correlate log events, and machine learning methods to evaluate the severity, priority and relation of log events and forecast the forthcoming critical faults of the monitored system. We have evaluated the prototyping implementation of the proposed system on a high performance computing cluster system and provided analysis with lessons learned.

Autonomous fault management of network and distributed systems is a challenging research problem and attracts many research activities. Solving this problem heavily depends on expertise knowledge and supporting tools for monitoring and detecting defects automatically. Recent research activities have focused on machine learning techniques that scrutinize system output data for mining abnormal events and detecting defects. This paper proposes a two-phase defect detection for network and distributed systems using log messages clustering and classification. The approach takes advantage of K-means clustering method to obtain abnormal messages and random forest method to detect the relationship of the abnormal messages and the existing defects. Several experiments have evaluated the performance of this approach using the log message data of Hadoop Distributed File System (HDFS) and the bug report data of Bug Tracking System (BTS). Evaluation results have disclosed some remarks with lessons learned.

Geometric modeling is a research based on computational geometry for processing the 3D objects in computer graphics or image processing. Rendering the graphical objects in medical applications is an interesting research that can help doctors have an exact decision in their diagnostic method. Application of geometric modeling in visualization of medical data is considered as a tool to support treatment. In the field of big data analysis and data science, the methods, tools and techniques are essential and important to analyze a massive amount of information and make data-driven decisions. These data are first analyzed and computed depending on the criteria and purpose of users. They are then visualized or simulated using visual elements like charts, graphs or maps to explore, understand their characteristics and data structure. Which method or solution is best for rendering and visualizing the analyzed data? That is still a question and even a challenge to the researchers. In this article, we apply geometric modeling to build a web-based application for visualizing a medical image dataset. This is an improved version of the selected paper in the previous work [1]. We first review several researches in the fields of computer graphics, images processing, geometric modeling and their applications; then we combine all of them and propose a method to develop our application for reconstructing the medical data objects. The input data are a Dicom (Digital Imaging and Communications in Medicine) dataset captured from CT (computed tomography) scanner or MRI (Magnetic Resonance Imaging). The output are 2D and 3D models that can be displayed and handled in different ways on the web interface. In comparison with other existing applications, our method shown the advantages of its using in practice.

Building 3D objects or reconstructing their surfaces from 3D point cloud data are researched activities in the field of geometric modeling and computer graphics. In the recent years, they are also studied and used in some fields such as: graph models and simulation; image processing or restoration of digital heritages. This article presents an improved method for restoring the shape of 3D point cloud surfaces. The method is a combination of creating a Bezier surface patch and computing tangent plane of 3D points to fill holes on a surface of 3D point clouds. This method is described as follows: at first, a boundary for each hole on the surface is identified. The holes are then filled by computing Bezier curves of surface patches to find missing points. After that, the holes are refined based on two steps (rough and elaborate) to adjust the inserted points and preserve the local curvature of the holes. The contribution of the proposed method has been shown in processing time and the novelty of combined computation in this method has preserved the initial shape of the surface

Auto-reconfiguration and response mechanisms in IP networks are challenges that lead to enforcing the required policies in such a dynamic environment. Software-Defined Network (SDN) provides an economic architecture solution for service providers by reducing provisioning time from weeks to minutes, facilitating centralized management, allowing programmability and promoting innovation. SDN achieves this by separating the two features inside network devices, such as routers and switches, namely the control plane from the data plane. The paper presents an implementation of a web-based management system for SDN controllers. This implementation allows network operators to control and supervise quickly the virtual network functions (VNF) provided by SDN networks. The implementation not only takes advantages of open source software including Flask and Ryu frameworks, OpenvSwitch and Mininet but also supports for virtual network functions. The system emulates a network topology with simple switches and links and also acts as a management center to query commands from Ryu network application using REST API in order to monitor changes in network traffic.

Bug tracking systems manage bug reports for assuring the quality of software products. A bug report (alsoreferred as trouble, problem, ticket or defect) contains several features for problem management and resolution purposes. Severity and priority are two essential features of a bug report that define the effect level and fixing order of the bug. Determining these features is challenging and depends heavily on human being, e.g., software developers or system operators, especially for assessing a large number of error and warning events occurring on software products or network services. This study first proposes a comparison of machine learning techniques for assessing severity and priority for software bug reports and then chooses an approach of using optimal decision trees, or random forest, for further investigation. This approach aims at constructing multiple decision trees based on the subsets of the existing bug dataset and features, and then selecting the best decision trees to assess the severity and priority of new bugs. The approach can be applied for detecting and forecasting faults in large, complex communication networks and distributed systems today. We have presented the applicability of random forest for bug report analysis and performed several experiments on software bug datasets obtained from open source bug tracking systems. Random forest yields an average accuracy score of 0.75 that can be sufficient for assisting system operators in determining these features. We have provided some analysis of the experimental results.

Occlusions become conspicuous as failure regions in video processing when unified over time because the contraventions of the restriction of brightness have accumulated and evolved in occluded regions. The accuracy at the boundaries of the moving objects is one of the challenging areas that required further exploration and research. This paper presents the work in process approach that can detect occlusion regions by using pixel-wise coherence, segment-wise confidence and interpolation technique. Our method can get the same result as usual methods by solving only one Partial Differential Equations (PDE) problem; it is superior to existing methods because it is faster and provides better coverage rates for occlusion regions than variation techniques when tested against a varied number of benchmark datasets. With these improved results, we can apply and extend our approach to a wider range of applications in computer vision, such as background subtraction, tracking, 3D reconstruction, video surveillance, video compression.

The 21st century has witnessed the rapid development of small and convenient mobile devices such as smartphones, tablets, game players, sensor nodes, etc. The rise of such mobile devices indicates the increase of colossal data transmission through the Internet and online services along with the challenges of data security. It is common to think of a solution to protect sensitive data from unauthorized users, and the most popular solution is to use encryption. While many research activities in functional encryption have widely been applied to network devices, computers, and applications, mobile devices still attract much attention to security issues due to the limitations of system resources, connectivity, data transmission and power consumption that malicious users can exploit to launch attacks. Especially, mobile devices have become a principal tool to share data on the Internet through online services, such as Facebook, Youtube, DropBox, Amazon, Online Games, etc. This paper presents a study of the Attribute-Based Encryption (ABE) scheme that exploits user attributes to build the secret key and the ciphertext. ABE encryption is specified by a set of attributes or a policy defining attributes that users possess. The paper also describes a few implementations of ABE applied in the cryptography community and the challenges of integrating ABE into real-world applications. Finally, the paper proposes an implementation of ABE for Android mobile devices. This implementation associated with the Kerberos protocol can be applied to secured data sharing applications. The Kerberos protocol aims at providing mutual authentication for the client server model. Experiments have evaluated the proposed ABE implementation on Android mobile devices along with the Kerberos system. The evaluation also includes ABE performance with discussions and lessons learned.

Traffic surveillance system (TSS) has seen great progress in the last several years. Many algorithms have been developed to cope with a wide range of scenarios such as overcast, sunny weather that created shadows, rainy days that result in mirror reflection on the road, or nighttime when low lighting conditions limit the visual range. However, in real-world applications, one of the most challenging problems is the scene determination in a highly dynamic outdoor environment. As also pointed out in recent survey, there have been limited studies on a mechanism for scene recognition and adapting appropriate algorithms for that scene. Therefore, this research presents a scene recognition algorithm for all-day surveillance. The proposed method detects and classifies outdoor surveillance scenes into four common types: overcast, clear sky, rain, and nighttime. The major contributions are to help diminish hand-operated adjustment and increase the speed of responding to the change of alfresco environment in the practical system. To obtain high reliable results, we combine the histogram features on RGB color space with the probabilistic model on CIE-Lab color space and input them into a feedforward neural network. Early experiments have suggested promising results on real-world video data.

Bug tracking systems manage bug reports for assuring the quality of software products. A bug report also referred as trouble, problem, ticket or defect contains several features for problem management and resolution purposes. Severity and priority are two essential features of a bug report that define the effect level and fixing order of the bug. Determining these features is challenging and depends heavily on human being, e.g., software developers or system operators, especially for assessing a large number of error and warning events occurring on software products or network services. This study proposes an approach of using random forest for assessing severity and priority for software bug reports automatically. This approach aims at constructing multiple decision trees based on the subsets of the existing bug dataset and features, and then selecting the best decision trees to assess the severity and priority of new bugs. The approach can be applied for detecting and forecasting faults in large, complex communication networks and distributed systems today. We have presented the applicability of random forest for bug report analysis and performed several experiments on software bug datasets obtained from open source bug tracking systems. Random forest yields an average accuracy score of 0.75 that can be sufficient for assisting system operators in determining these features. We have provided some analysis of the experimental results.

The Uber-based applications have recently created a new business model: a taxi company without any car, a tutor company without any tutor, or a hotel without any room. These applications coordinate mobile computing and peer-to-peer technology to facilitate the peer-to-peer provision of services. This paper presents a study of Uber-based applications. The paper first explains the driving force of mobile computing and peer-to-peer technology that exploit direct communications between mobile applications for services. It then describes a common application framework with the system architecture and prevailing components. We use virtual healthcare and software outsourcing case studies to demonstrate the prototyping systems with functions and evaluate service availability and performance.

Monitoring events in communication and computing systems becomes more and more challenging due to the increasing complexity and diversity of these systems. Several supporting tools have been created to assist system administrators in monitoring an enormous number of events daily. The main function of these tools is to filter as many as possible events and present highly suspected events to the administrators for fault analysis, detection and report. While these suspected events appear regularly on large and complex systems, such as cloud computing systems, analyzing them consumes much time and effort. In this study, we propose an approach for evaluating the severity level of events using a classification decision tree. The approach exploits existing fault datasets and features, such as bug reports and log events to construct a decision tree that can be used to classify the severity level of other events. The administrators refer to the result of classification to determine proper actions for the suspected events with a high severity level. We have implemented and experimented the approach for various bug report and log event datasets. The experimental results reveal that the accuracy of classifying severity levels by using the decision trees is above 80%, and some detailed analyses are also provided.

In the field of traffic monitoring systems, shadows are the main causes of errors in computer vision-based vehicle detection and classification. A great number of research have been carried out to detect and remove shadows. However, these research works only focused on solving shadow problems in daytime traffic scenes. Up to now, far too little attention has been paid to the problem caused by vehicles’ reflections in rainy conditions. Unlike shadows in the daytime, which are homogeneous gray shades, reflection shadows are inhomogeneous regions of different colors. This characteristic makes reflections harder to detect and remove. Therefore, in this paper, we aim to develop a reflection detection and removal method from single images or video. Reflections are detected by determining a combination of L and B channels from LAB color space and H channel from HSV color space. The reflection removal method is performed by determining the optimal intensity of reflected areas so that they match with neighbor regions. The advantage of our method is that all reflected areas are removed without affecting vehicles’ textures or details.

Software defined network separates data and control planes that facilitate network management functions, especially enabling programmable network control functions. Event monitoring is a fault management function involved in collecting and filtering event notification messages from network devices. This study presents an approach of distributed event monitoring for software defined network. Monitoring events usually deals with a large amount of event log data, log collecting and filtering processes thus require a high degree of automation and efficiency. This approach takes advantage of the OpenFlow and syslog protocols to collect and store log events obtained from network devices on a syslog server. It also uses the adaptive semantic filtering method to filter and present non-trivial events for system administrators to take further actions. We have evaluated this approach on a network simulation platform and provided some log collection and filtering results with analysis.