SRv6 can provide hybrid cooperation between a centralized network controller and network nodes. IPv6 routers maintain multi-hop ECMP-aware segments, whereas the controller establishes a source-routed path through the network. Since the state of the flow is defined at the ingress to the network and then is contained in a specific packet header, called Segment Routing Header (SRH), the importance of such a header itself is vital. Motivated by the need to study and investigate this technology, this paper discusses some security-related issues of Segment Routing. A SRv6 capable experimental testbed is built and detailed. Finally, an experimental test campaign is performed and results are evaluated and discussed
User-Equipments (UEs) capable of working with cloud computing have grown exponentially in recent years, leading to a significant increase in the amount of data production. Moreover, upcoming Internet-of-Things (IoT) applications such as virtual and augmented reality, video streaming, intelligent transportation, and healthcare will require low latency, communications, and processing. Edge computing is a revolutionary criterion in which dispersed edge nodes supply resources near end devices because of the limited resources available on UEs. Rather than transmitting massive amounts of data to the cloud, edge nodes could filter, analyze, and process the data they receive using local resources. Mobile Edge Computing (MEC), in particular, when user mobility is considered, has the potential to significantly reduce processing delays and network traffic between UEs and servers. This research demonstrated a novel technique for migration that minimizes delay and downtime by utilizing edge computing. Our proposed method syncs more frequently than the pre-copy method which is the most used migration method that synchronizes (sync) the source and destination only based on multiple rounds. When compared to established migration methodologies, our results indicate that our mechanism has less latency, downtime, migration time, and packet loss. These results allow delay-sensitive applications that require ultra-low latency to function smoothly during migration.
The new 5G mobile network promises to enhance existing services or include new ones to address key challenges presented by smart city stakeholders (citizens, municipalities, politics, industries, architects, etc.) to improve system implementations. These challenges cover various smart city fields such as transportation, environmental monitoring, healthcare, industrial automation, smart grid, etc. Thus, the main objective of 5G functionalities is to provide solutions to the various identified needs, which are defined as constraints and requirements. Therefore, three categories of 5G-based use cases have been defined: Enhanced Mobile Broadband (eMBB), Massive Machine Type Communications (mMTC), and Ultra-reliable and Low Latency Communications (uRLLC). Each group involves a set of use cases and characterized by specific technical features that address the corresponding needs. However, accurate and real-time positioning information is a vital requirement common to all three categories, but the degree of performance varies across scenarios and descriptions. Therefore, this work presents a summary of existing positioning technologies crossed with wireless technologies and smart city use cases to highlight the potential that will add accurate and real-time positioning to 5G capabilities. 5G promises decimeter accuracy in some critical use cases.
Although awareness about the importance of injury prevention has been increasing among Japanese people, preventable injuries remain the third leading cause of death in children aged 0–14 years, and prevention of these injuries is critically important in terms of childhood health. To identify dangerous situations for children and provide preventive measures to avoid such situations, this paper proposes an effective method, called “Empowering Reality (ER)”, that integrates knowledge graphs with object detection to enable lecturers to educate caregivers on preventing unintentional childhood injuries while communicating with caregivers using augmented reality technology. The proposed ER system consists of knowledge graphs for explaining dangerous situations, an online video capture part, and a situation recognition part. This paper describes the major advantages of knowledge graphs that consider not only the relationship between objects and injuries, but also dangerous layouts with the help of “inclusion” and “collocation” features. The feasibility and effectiveness of the system were evaluated through tests among caregivers, including 11 parents and six teachers from three nursery schools. This system allows lecturers to conduct in-situ suggestions about specific preventive measures adapted to the home or nursery school environment via online learning
Towards realising autonomous UAVs, this paper investigates one of the fundamental autonomous flying research problems, i.e., the ability of a vehicle to control its flying behaviour autonomously, without reliance on external infrastructure like Instrument Landing Systems or GPS. In this paper we experiment with a physical UAV prototype with embedded intelligent control capabilities, utilising a Long Short Term Memory (LSTM) neural network, in order to learn lift-off control sequences using self-training. The initial results are promising and show potential for embedding LSTMs in the control systems of autonomous UAVs.
Currently, in a modern world where people are getting busier, it is hard for them to take time to work-out or exercise regularly by themselves. People have been using the gym as a place to make their body fit and it is right as Health and Immunity are an important part of a person’s life and everyone would always like to be fit and healthy. To achieve that level requires motivation and discipline. And lack of motivation results in quitting the gym in a very short span. Now Generally in Gym, there are only a few that keep track of everything, others we have a huge marketplace for the people who joined but quit soon since doing exercise doesn’t give you short-term results. The changes in your body start appearing after months. The slightest changes and every other important aspect can be hard to keep track of them altogether, now using smart gym approach, the slightest change can be measured, using IoT and later one can curate all the training regiments, diets and exercises properly according to it. We propose an overall IoT-based-system to monitor the user's Health and Fitness Records in an effective way of using a database management system. It plans to collect data from the machines when the user uses it, keep track of its workouts and diet intake, with a gym social media that would help to maintain a competitive environment, also including management of membership, payment, trainers, and employees. This would result in a whole new way of looking towards the gym
The aim of this paper is to investigate the use of data reduction techniques using distances and areas for monitoring of sensors using process mining approaches. When sensors are used in industrial cases, the real time accumulation of timestamped data tends to pose a problem of storage, processing and analyzing their values in the optimal way possible. Here, the paper tends to present an application of monitoring of sensor signals using reduced parameters from the acquired timestamped values. Each observation is dissected into packets and their attributes such as areas under the curves and successive distances are calculated. The combination of these attributes present real time monitoring scenario for which a blueprint process model can be constructed. This, in turn, helps in identifying signal variations during run-time of the sensor without advanced analysis.
Modern technologies like digital spaces, intelligent transportation, and digital operations are faced with technical challenges related to data capturing, data processing, data storage, data security, communication, etc. Ensuring security and reliable message propagation among autonomous vehicles is a major challenging task. Establishing appropriate environment for developing vehicular-based solutions by the vehicle manufactures increases their estimated cost and time. Hence, to minimize this problem, this proposal aims to introduce Service Oriented Architecture (SOA) principles, to access trust based solutions as simple cloud based services by the clients. We propose a framework to integrate three mandatory vehicular applications namely trust estimation, secured message dissemination, and routing as cloud-based microservices. We also propose an innovative CipherText Policy Attribute Based Encryption (CP-ABE) algorithm to ensure confidentiality of data by an access control system in highly dynamic and automated network. The services are deployed as Docker images using advanced concepts of Dockers and Containers. Dockers coordinate the orchestration of multiple tasks related to the proposed microservices and help to implement the services in cross-platform environments. These services can be implemented in both autonomous and manual vehicular systems. The service providers can charge the clients based on the usage of the services. A detailed experimental analysis is accomplished to evaluate the performance of the proposed micro services in cross platform environments; further, an extensive simulation is performed to assess the individual performance of the proposed vehicular applications.
Mobile apps are becoming increasingly important in everyone's daily life. The success of an app is linked to high user acceptance. Therefore, it is necessary to capture users' expectations, needs, and problems regarding an app in any situation. By continuously capturing and analyzing user feedback, developers can evaluate the level of user acceptance. There are various feedback channels, such as app stores, social networks, and within the app, which can be used to capture user feedback. As we already have experience with feedback from app stores and social networks, we wanted to investigate inapp feedback approaches and thus conducted a mapping study to understand the state of the art of these approaches.We analyzed 36 publications and derived requirements for in-app feedback tools. Based on that, we defined requirements for an in-app feedback tool to describe its prototypical realization. Then we performed an evaluation regarding user acceptance of our tool with 33 participants. The evaluation showed a high rate of acceptance for the tool among the participants. The results also highlighted improvement areas for our tool, such as optimizing the rate of requests for feedback. We plan to address these aspects in future work and to continue improving our tool.
Networks are pervasive in computer science and in real world applications. It is often useful to leverage distinctive node features to regroup such data in clusters, by making use of a single representative node per cluster. Such contracted graphs can help identify features of the original networks that were not visible before. As an example, we can identify contiguous nodes having the same discrete property in a social network. Contracting a graph allows a more scalable analysis of the interactions and structure of the network nodes. This paper delves into the problem of contracting possibly large colored networks into smaller and more easily manageable representatives. It also describes a simple but effective algorithm to perform this task. Extended performance plots are given for a range of graphs and results are detailed and discussed with the aim of providing useful use cases and application scenarios for the approach
The integrated impact of computing techniques and resources with big-data processing transforms human lifestyles by providing quality services ranging from healthcare to smart homes and effective interactions. However, many healthcare systems fail to consider patient emergencies and cannot provide a customized resource service. Cognitive computing is a requisite technology to create these intelligent systems based on artificial intelligence algorithms. This paper presents technologies for personalized healthcare services through cognitive computing. This paper investigates cognitive computing developments from discovering knowledge, cognitive science, and big-data analytics at the onset. Then, the system architecture for a cognitive computing system is given. Furthermore, this paper presents the technologies for cognitive computing healthcare improvement opportunities and their challenges. Finally, this paper discusses the representative intelligent systems of cognitive computing, including medical, robotic, and cognitive-communication systems.
Plans to mass evacuate visitors in an exhibition center in the case of emergency situations are critical for public safety and disaster management. Efficient crowd evacuation during mass gatherings has been an active research area during the past years. In this paper, we consider the challenging problem of finding in near real-time the most efficient and safest evacuation pathways in a multi-exit exhibition center while the fire hazard spreads. We first propose a system composed of sensor nodes to collect pertinent safety data associated with the changing environmental conditions. We then present a spontaneous dynamic evacuation system that considers the changing conditions in the risks associated with each hallway segment in terms of walking distance, heat, two major asphyxiant fire gases and crowd congestion. Our IoT-based system activates smart panels placed at major junctions of the hallways to visually guide evacuees towards the safest escape direction under the existing circumstances. The proposed algorithms aim to minimize the total evacuation time of all evacuees, while circumventing congested and perilous aisles, balancing traffic loads, and guaranteeing high scalability and reasonable computational efficiency. This work can pave the way towards the development of nextgeneration smart exhibition centers, where crowd safety is among the top priorities
Computing the convex hull of a set of points is a fundamental issue in many fields, including geometric computing, computer graphics, and computer vision. This problem is computationally challenging, especially when the number of points is past the millions. In this paper, we propose a fast filtering technique that reduces the computational cost for computing a convex hull for a large set of points. The proposed method preprocesses the input set and filters all points inside a four-vertex polygon. The experimental results showed the proposed filtering approach achieved a speedup of up to 77 and 12 times faster than the standard Graham scan and Jarvis march algorithms, respectively.
More freedom, more flexibility, and reduced travel time for knowledge workers are just a few advantages of new work models, which have been discussed for several years now. Moreover, the problem of rural depopulation can be addressed by this concept. In the research project “Digital Teams”, we aim to develop a digital open-source platform to support and optimize the digital work environment for distributed teams in rural areas, especially in the knowledge work context. In this article, we focus on the research and design aspects of the project. We provide insights on how we have used the design thinking approach for our research and the development of the UX- and UI-design concepts. We are focusing on an ecosystem concept, which provides all relevant services for knowledge workers in their daily work life, rather than focusing on a specific remote collaboration purpose. We present initial evaluation results, which tend to be positive and give an outlook on future work.
The route followed by a refugees’ group towards its destination can, in many cases, be regarded as the reference point around which the collective memory of such a group of people is intertwined. Such a route enriches people's memories with common experiences, targeting places and interactions among refugees and locals and may affect the collective memory of such people positively or negatively. A crucial point in the modern paradigm of smart cities is the quality of life. To achieve quality of life for its citizens a smart city should establish ways to reduce alienation among the different groups that constitute the city's palimpsest. Understanding the different cultural identities and improvement of social cohesion between different people groups is one of the basic vehicles towards this goal. In this paper, we attempt to give a first answer to such problems proposing and implementing specific services in the context of a crowdsourcing system for collective memory management using interactive maps. We demonstrate a basic usage scenario to show the strength of the implemented services, along with a two-step evaluation showing positive results.
In the industrial revolution that followed, the aim of the industry is not only to improve and meet its urgent needs, but also to improve the standard of living of society and make life easier for consumers. Therefore, economic growth must always be closely linked to the industrial revolution. The medical industry, energy conservation and, in particular, production technologies will be transformed through new value chain models. Globalization, urbanization, vital changes and the energy transformation are all shifting forces that assess the dynamics of technology to better identify solutions in the moving world. In recent years, successive revolutions have made remarkable contributions to a person's quality of life, safety, industrial economy, comfort and health. This work aims to improve the energy consumption of the building. To achieve this goal, digital twins were created to faithfully reflect the behavior and characteristics of future or current buildings. To replicate copies of future or existing buildings, we chose to use Autodesk REVIT solution to meet some limitations. These have a great influence on the energy behavior of the building.
The new generation of networking environments such as the internet of things (IoT), cloud computing, etc. is emerging and releases new prospects to traditional information systems by merging new technologies and services for seamless access to information sources at anytime and anywhere. Concurrently, this emergence opens new threats to information security and new challenges to controlling access to resources. To ensure security, several techniques have been employed, and access control (AC) is one of the essential security requirements especially for recent networking environments. Various authentication and AC methods are proposed to enforce AC policy and to prevent any unauthorized access to logical/physical assets. The continuous technology upgrades and the diversity of AC models force the need to find AC metamodels with a higher level of abstraction that serves as a unifying framework for specifying any AC policy. AC metamodels are proposed to encompass AC features and are used to derive various instances of AC models and methods. In this paper we review the proposed AC metamodels and their implementation scenarios, we analyze them, their objectives, their limitations, and present current research issues and open questions that still need to be addressed.
ackers are increasingly launching phishing attacks via SMS and social media. Games and dating apps introduce yet another attack vector. However, current deep learning-based phishing detection applications do not apply to mobile devices due to the computational burden. We propose a lightweight phishing detection algorithm that distinguishes phishing from legitimate websites solely from URLs to be used in mobile devices. As a baseline performance, we apply Artificial Neural Networks (ANNs) to URL-based and HTML-based website features. A model search results in 15 ANN models with accuracies >96%, comparable to state-of-the-art approaches. Next, we test the performance of deep ANNs on URLbased features only; however, all models perform poorly with the highest accuracy of 86.2%, indicating that URL-based features alone are not adequate to detect phishing websites even with deep ANNs. Since language transformers learn to represent context-dependent text sequences, we hypothesize that they will be able to learn directly from the text in URLs to distinguish between legitimate and malicious websites. We apply three state-of-the-art deep transformers (BERT, ELECTRA, and RoBERTa) for phishing detection. Testing custom and standard vocabularies, we find that pre-trained transformers available for immediate use (with fine-tuning) outperform the model trained with the custom URL-based vocabulary. In addition, we test a thinner BERT transformer which is suitable for lightweight devices like mobiles, called MobileBERT. Our results emphasize that evaluation metrics of this model are competitive to other models in this study, yet the testing time is significantly less, making this model a choice for embedding phishing detection algorithms in mobile phones. Using pre-trained transformers to predict phishing websites from only URLs has five advantages: 1) requires little training time (230 to 320 s), 2) is more easily updatable than feature-based approaches because no pre-processing of URLs is required, 3) is safer to use because phishing websites can be predicted without physically visiting the malicious sites, 4) is easily deployable for real-time detection and is applicable to run on mobile devices, and 5) using a mobile specific transformer yields comparable performance and predicts 3 times faster than the other transformer models in this study.
Indoor localization is of great importance to several fields such as healthcare and asset tracking. However, many factors (e.g., multipath propagations) impact the quality of signals which are used to perform localizations. As a consequence, the precision and accuracy of the computed locations are heavily influenced. Therefore, the methodologies to compute indoor locations always need continuous refinements in terms of those metrics including the time complexity. For the last metric, It impacts the performance of mobile devices due to their limited resources. To address these challenges, a new set of fingerprinting algorithms was presented in this paper called Fingerprinting Line-Based Nearest Neighbour. This set shifts grid points potentially towards targets via a deterministic percentage. The running time of the set is upper bounded. Moreover, this paper presents the following: 1) an upper bound in terms of distance errors for the proposed algorithms, and 2) based on real experiments, the new algorithms (e.g., 90% shifting) improved the accuracy and precision, and had lower distance errors probabilities compared to those for the nearest neighbour-based algorithms (e.g., by 106% and 76%, respectively).
