RonPub -- Research Online Publishing
Not a member yet
199 research outputs found
Sort by
MASCARA (ModulAr Semantic CAching fRAmework) towards FPGA Acceleration for IoT Security Monitoring
Full text linkWith the explosive growth of the Internet Of Things (IOTs), emergency security monitoring becomes essential to efficiently manage an enormous amount of information from heterogeneous systems. In concern of increasing the performance for the sequence of online queries on long-term historical data, query caching with semantic organization, called Semantic Query Caching or Semantic Caching (SC), can play a vital role. SC is implemented mostly in software perspective without providing a generic description of modules or cache services in the given context. Hardware acceleration with FPGA opens new research directions to achieve better performance for SC. Hence, our work aims to propose a flexible, adaptable, and tunable ModulAr Semantic CAching fRAmework (MASCARA) towards FPGA acceleration for fast and accurate massive logs processing applications
ParkingJSON: An Open Standard Format for Parking Data in Smart Cities
Full text linkData marketplaces and data management platforms offer a viable solution to build large city-scale Internet of Things (IoT) applications. Contemporary data marketplaces and data management platforms for smart cities such as Intelligent IoT Integrator (I3), Cisco Kinetic, Terbine, and Streamr present a middleware platform to help the data owners to provide their data to the application developers. However, such platforms suffer from adoption issues because of the interoperability concerns that stem from heterogeneous data formats. On the one hand, the IoT devices and the software used by the device owners follow either a custom data standard or a proprietary industrial standard. On the other hand, the application developers consuming data from multiple device owners expect the data to follow one common standard to process the data without developing custom software for each data feed. Therefore, a common data standard is desired to enable interoperable data exchange through data marketplace and data management platforms while promoting adoption. We present our experiences from developing a city-scale real-time parking application for a smart city. We also introduce ParkingJSON, a new open standard format for parking data in smart cities, which could help the parking data providers to cover all types of parking infrastructures through a single JSON schema. To the best of our knowledge, this is the first parking data standard proposed that a) covers a wide range of parking spaces and structures, b) integrates spatial information, and c) provides support for data integrity and authenticity
NebulaStream: Complex Analytics Beyond the Cloud
Full text linkThe arising Internet of Things (IoT) will require significant changes to current stream processing engines (SPEs) to enable large-scale IoT applications. In this paper, we present challenges and opportunities for an IoT data management system to enable complex analytics beyond the cloud. As one of the most important upcoming IoT applications, we focus on the vision of a smart city. The goal of this paper is to bridge the gap between the requirements of upcoming IoT applications and the supported features of an IoT data management system. To this end, we outline how state-of-the-art SPEs have to change to exploit the new capabilities of the IoT and showcase how we tackle IoT challenges in our own system, NebulaStream. This paper lays the foundation for a new type of systems that leverages the IoT to enable large-scale applications over millions of IoT devices in highly dynamic and geo-distributed environments
Can You Hear Me? A Metric for Link Asymmetry
Full text linkThe Internet of Things is a networking paradigm aiming to provide computing pervasiveness to our everyday lives. A key component to the Internet of Things is low power networks that gather information from the environment. Low power networks are prone to asymmetric and unidirectional links. Measuring the level of asymmetry and understanding its sources are key steps to successfully deploying sensor networks and the Internet of Things. Our first contribution is a new metric to assess link asymmetry, one which takes into account the instantaneous delivery success probability. Next, we study the influence of four factors on link asymmetry in light of our asymmetry metric, namely, relative distance, output power, relative position, and hardware heterogeneity. With our unique method, we show that all four factors impact link asymmetry
Multi-Game Code-Duel for Learning Programming Languages
Full text linkSoftware developers compose computer instructions following the rules defined in programming languages for the purpose of automatic information processing. However, different programming languages have different syntax and semantic rules, and support different programming paradigms and design patterns. Learning a programming language needs many efforts and much practicing in order to master the rules and apply the patterns. Leaning multiple programming languages at the same time, of course, needs more efforts. In this work we develop the concept of multi-game and an e-learning platform called "Multi-Game Platform for Code-Duels" for learning multiple programming languages easily and efficiently. A multi-game is a video game, which consists of several mini-games. Dividing a big game into mini-games reduces the development efforts and implementation complexity. "Builders" is a multi-game developed in our platform consisting of three mini-games. Each mini-game can be solved by implementing a program by learners using different languages. Using our multi-game platform, each mini-game of Builders can be developed easily and played independently of the other mini-games. Finally, a user evaluation over our multi-game platform is performed, where users rate our multi-game approach and platform for learning programming languages very positively
Experimentation and Analysis of Ensemble Deep Learning in IoT Applications
Full text linkThis paper presents an experimental study of Ensemble Deep Learning (DL) techniques for the analysis of time series data on IoT devices. We have shown in our earlier work that DL demonstrates superior performance compared to traditional machine learning techniques on fall detection applications due to the fact that important features in time series data can be learned and need not be determined manually by the domain expert. However, DL networks generally require large datasets for training. In the health care domain, such as the real-time smartwatch-based fall detection, there are no publicly available large annotated datasets that can be used for training, due to the nature of the problem (i.e. a fall is not a common event). Moreover, fall data is also inherently noisy since motions generated by the wrist-worn smartwatch can be mistaken for a fall. This paper explores combing DL (Recurrent Neural Network) with ensemble techniques (Stacking and AdaBoosting) using a fall detection application as a case study. We conducted a series of experiments using two different datasets of simulated falls for training various ensemble models. Our results show that an ensemble of deep learning models combined by the stacking ensemble technique, outperforms a single deep learning model trained on the same data samples, and thus, may be better suited for small-size datasets
Integrating a Smart City Testbed into a Large-Scale Heterogeneous Federation of Future Internet Experimentation Facilities: the SmartSantander Approach
Full text linkFor some years already, there has been a plethora of research initiatives throughout the world that have deployed diverse experimentation facilities for Future Internet technologies research and development. While access to these testbeds has been sometimes restricted to the specific research community supporting them, opening them to different communities can not only help those infrastructures to achieve a wider impact, but also to better identify new possibilities based on novel considerations brought by those external users. On top of the individual testbeds, supporting experiments that employs several of them in a combined and seamless fashion has been one of the main objectives of different transcontinental research initiatives, such as FIRE in Europe or GENI in United States. In particular, Fed4FIRE project and its continuation, Fed4FIRE+, have emerged as "best-in-town" projects to federate heterogeneous experimentation platforms. This paper presents the most relevant aspects of the integration of a large scale testbed on the IoT domain within the Fed4FIRE+ federation. It revolves around the adaptation carried out on the SmartSantander smart city testbed. Additionally, the paper offers an overview of the different federation models that Fed4FIRE+ proposes to testbed owners in order to provide a complete view of the involved technologies. The paper is also presenting a survey of how several specific research platforms from different experimentation domains have fulfilled the federation task following Fed4FIRE+ concepts
Towards a Large Scale IoT through Partnership, Incentive, and Services: A Vision, Architecture, and Future Directions
Full text linkInternet of Things applications has been deployed and managed in a small to a medium scale deployments in industries and small segments of cities in the last decade. These real-world deployments not only helped the researchers and application developers to create protocols, standards, and frameworks but also helped them understand the challenges associated with the maintenance and management of IoT deployments in all kinds of operational environments. Despite the technological advancements and the deployment experiences, the technology failed to create a notable momentum towards large scale IoT applications involving thousands of IoT devices. We argue the reasons behind the lack of large scale deployments and the limitations of contemporary IoT deployment model. In addition, we present an approach involving multiple stakeholders as a means to scale IoT applications to hundreds of devices. Besides, we argue that the partnership, incentive mechanisms, privacy, and security frameworks are the critical factors for large scale IoT deployments of the future
Towards an Inclusive Definition and Framework Development for M-Learning
Full text linkMobile learning has changed the course of learning in higher and tertiary education. However, there are still mixed views on the inclusive definition and best usable frameworks for implementing mobile learning in formal education system. Hence, the question, which has been posed but not been explicitly answered by researchers, is: What is the correct view of mobile learning? This question has left so many researchers mystified but the answer lies in the way in which mobile learning is defined. How then should mobile learning be defined? This article serves to propose an inclusive definition that can be used to guide the development of mobile learning systems in formal education. In addition to the guide, this paper proposes a framework for usage and implementing multimedia mobile e-learning
Adding Semantics to Enrich Public Transport and Accessibility Data from the Web
Full text linkWeb technologies and open data practices have now begun to promote new issues and services addressed to both final and specialized users. The smart cities initiative has also introduced new trends and ideas to offer to the public, one of which is the challenge of a more inclusive society that will provide the same opportunities for all. One of the major areas that could benefit from these new initiatives is public transport by, for example, providing open and accessible datasets, which include information by and about people with special needs. In this sense, the Google Transit Feed Specification (GTFS) defines a format to describe public transportation and associated geographic information. It includes details regarding accessibility and what people with special needs might require to get around using public transport. We are, however, of the opinion that this specification has a low granularity and is not sufficient, since it only takes into account only mobility needs. As suggestions for improvement, we propose to enrich GTFS data by combining public transport data from multiple Web sources with semantic metadata techniques. Those data are stored in a public semantic dataset. To define this dataset, we propose a systematic method to extract data from different sources and integrate them. This method is applied to obtain data about the metro system from the website of Metro Madrid and GTFS. Relevant SPARQL queries and two applications are developed to evaluate the usefulness of the dataset obtained