RonPub -- Research Online Publishing
Not a member yet
199 research outputs found
Sort by
Data Lifetime Estimation in a Multicast-Based CoAP Proxy
Full text linkIn this work we consider kernel-based record lifetime estimation in a proactive Internet of Things (IoT) proxy with multicast based cache management. Multicast refreshment requests were based on lifetime expiration for a predefined number of records. To reduce the traffic volume in the IoT domain, we assume that only nodes where the observed physical variable has changed its value will respond to the multicast request. For estimating the data lifetime at the proxy, we use Gaussian kernels, assuming that the intrinsic data lifetime probability distribution was taken from Erlang-k family of sub-exponential distributions. In this setup, we consider that the proxy connects to the IoT domain using an IEEE 802.15.4-compatible wireless network. Results indicate that narrow and symmetrical lifetime probability distributions require more frequent multicasting refreshments compared to wider and asymmetric ones. This increases traffic intensity and energy consumption in IoT domain. We quantify finding with numerical results
NextGen Multi-Model Databases in Semantic Big Data Architectures
Full text linkWhen semantic big data is managed in commercial settings, with time, the need may arise to integrate and interlink records from various data sources. In this vision paper, we discuss the potential of a new generation of multi-model database systems as data backends in such settings. Discussing a specific example scenario, we show how this family of database systems allows for agile and flexible schema management. We also identify open research challenges in generating sound triple-views from data stored in interlinked models, as a basis for SPARQL querying. We then conclude with a general overview of multi-model data management systems, to provide a wider scope of the problem domain
Leveraging Application Development for the Internet of Mobile Things
Full text linkSo far, most of research and development for the Internet of Things has been focused at systems where the smart objects, WPAN beacons, sensors, and actuators are mainly stationary and associated with a fixed location (such as appliances in a home or office, an energy meter for a building), and are not capable of handling unrestricted/arbitrary forms of mobility. However, our current lifestyle and economy are increasingly mobile, as people, vehicles, and goods move independently in public and private areas (e.g., automated logistics, retail). Therefore, we are witnessing an increasing need to support Machine to Machine (M2M) communication, data collection, and processing and actuation control for mobile smart things, establishing what is called the Internet of Mobile Things (IoMT). Examples of mobile smart things that fit in the definition of IoMT include Unmanned Aerial Vehicles (UAVs), all sorts of human-crewed vehicles (e.g., cars, buses), and even people with wearable devices such as smart watches or fitness and health monitoring devices. Among these mobile IoT applications, there are several that only require occasional data probes from a mobile sensor, or need to control a smart device only in some specific conditions, or context, such as only when any user is in the ambient. While IoT systems still lack some general programming concepts and abstractions, this is even more so for IoMT. This paper discusses the definition and implementation of suitable programming concepts for mobile smart things - given several examples and scenarios of mobility-specific sensoring and actuation control, both regarding smart things individually, or in terms of collective smart things behaviors. We then show a proposal of programming constructs and language, and show how we will implement an IoMT application programming model, namely OBSACT, on the top of our current middleware ContextNet
Understanding the Performance of Software Defined Wireless Sensor Networks under Denial of Service Attack
Full text linkWireless sensor networks (WSN) are formed from restricted devices and are known to be vulnerable to denial of service (DoS) security attacks. In parallel, software-defined networking has been identified as a solution for many WSN challenges with respect to flexibility and reuse. Conversely, the SDN control plane centralization may bring about new security threats and vulnerabilities. In this work, we perform a traffic analysis of software-defined WSN (SDWSN) in order to gain understanding of the network's performance when it is under certain types of DoS attacks. In particular, we consider three different DoS scenarios of increasing aggressiveness: (i) false flow requests DoS, (ii) false data flow forwarding DoS, and, (iii) false neighbor information passing DoS. Our simulation results for the latter two types of attack showed significant changes both in the average value and the variance of the delivery rate and the overall overhead. These results demonstrate that it is possible to identify when a SDWSN is under a particular type of DoS, by monitoring the respective quantities
Online Replication Strategies for Distributed Data Stores
Full text linkThe rate at which data is produced at the network edge, e.g., collected from sensors and Internet of Things (IoT) devices, will soon exceed the storage and processing capabilities of a single system and the capacity of the network. Thus, data will need to be collected and preprocessed in distributed data stores - as part of a distributed database - at the network edge. Yet, even in this setup, the transfer of query results will incur prohibitive costs. To further reduce the data transfers, patterns in the workloads must be exploited. Particularly in IoT scenarios, we expect data access to be highly skewed. Most data will be store-only, while a fraction will be popular. Here, the replication of popular, raw data, as opposed to the shipment of partially redundant query results, can reduce the volume of data transfers over the network. In this paper, we design online strategies to decide between replicating data from data stores or forwarding the queries and retrieving their results. Our insight is that by profiling access patterns of the data we can lower the data transfer cost and the corresponding response times. We evaluate the benefit of our strategies using two real-world datasets
Editorial of the 2019 Workshop on Very Large Internet of Things (VLIoT)
Full text linkWe are proud of presenting the outcome of this third edition of the "Very Large Internet of Things" (VLIoT) workshop, which was held in Los Angeles (USA) in August 2019, in conjunction with the 45th International Conference on Very Large Data Bases (VLDB). Following the success path of the two previous workshop editions - in Munich (2017) and in Rio de Janeiro (2018) - VLIoT 2019 kept its tradition to be a vivid and high-quality technical forum for researchers and practitioners working with Internet of Things to share their experiences, visions and latest findings, most of them regarding the design, implementation, deployment and management of IoT systems at very large and scale. This editorial of the special issue introduces and introduces all papers presented at the workshop
Quasi-Convex Scoring Functions in Branch-and-Bound Ranked Search
Full text linkFor answering top-k queries in which attributes are aggregated to a scalar value for defining a ranking, usually the well-known branch-and-bound principle can be used for efficient query answering. Standard algorithms (e.g., Branch-and-Bound Ranked Search, BRS for short) require scoring functions to be monotone, such that a top-k ranking can be computed in sublinear time in the average case. If monotonicity cannot be guaranteed, efficient query answering algorithms are not known. To make branch-and-bound effective with descending or ascending rankings (maximum top-k or minimum top-k queries, respectively), BRS must be able to identify bounds for exploring search partitions, and only for monotonic ranking functions this is trivial. In this paper, we investigate the class of quasi-convex functions used for scoring objects, and we examine how bounds for exploring data partitions can correctly and efficiently be computed for quasi-convex functions in BRS for maximum top-k queries. Given that quasi-convex scoring functions can usefully be employed for ranking objects in a variety of applications, the mathematical findings presented in this paper are indeed significant for practical top-k query answering
Energy Savings in Very Large Cloud-IoT Systems
Full text linkOpposite to the original cloudlet approach in which an edge is utilized to bring the cloud and its benefits closer to the applications, in cloud- and edge-connected IoT systems where the applications are deployed and run in the cloud, we exploit the edge somewhat differently, either by bringing the physical world and its data up closer to the cloud or by caching parts of the applications down closer to the physical world. Aggressive optimizations seeking substantial IoT energy savings are needed to maintain the scalability of large-scale IoT deployments and to stay within cloud cost constraints (avoiding costly elasticity when working with a budget limit). In this paper, we present a novel optimization approach that relies on the simple principle of minimizing all movements: movements of data from the IoT up to the Edge and Cloud, and movements of application fragments from the cloud down to the edge and the IoT itself. Our approach is novel in that it involves and utilizes the dynamic characteristics and variability of both the data and applications simultaneously. Another novelty of our approach is the definition and use of "sentience-efficiency" as a precursor to "energy-efficiency" for achieving truly aggressive savings in energy. We present our bi-directional optimization approach and its implementation in terms of algorithms within an architecture we name the cloud-edge-beneath architecture (CEB). We present a performance evaluation study to measure the impact of our optimization approach on energy saving
Code Generation for Big Data Processing in the Web using WebAssembly
Full text linkTraditional clusters for cloud computing are quite hard to configure and setup, and the number of cluster nodes is limited by the available hardware in the cluster. We hence envision the concept of a Browser Cloud: One just has to visit with his/her web browser a certain webpage in order to connect his/her computer to the Browser Cloud. In this way the setup of the Browser Cloud is much easier than those of traditional clouds. Furthermore, the Browser Cloud has a much larger number of potential nodes, as any computer running a browser may connect to and be integrated in the Browser Cloud. New challenges arise when setting up a cloud by web browsers: Data is processed within the browser, which requires to use the technologies offered by the browser for this purpose. The typically used JavaScript runtime environment may be too slow, because JavaScript is an interpreted language. Hence we investigate the possibilities for computing the work-intensive part of the query processing inside a virtual machine of the web browser. The technology WebAssemby for virtual machines is recently supported by all major browsers and promises high speedups in comparison with JavaScript. Recent approaches to efficient Big Data processing generate code for the data processing steps of queries. To run the generated code in a WebAssembly virtual machine, an online compiler is needed to generate the WebAssembly bytecode from the generated code. Hence our main contribution is an online compiler to WebAssembly bytecode especially developed to run in the web browser and for Big Data processing based on code generation of the processing steps. In our experiments, the runtimes of Big Data processing using JavaScript is compared with running WebAssembly technologies in the major web browsers
Data-Centric Resource Management in Edge-Cloud Systems for the IoT
Full text linkA major challenge in emergent scenarios such as the Cloud-assisted Internet of Things is efficiently managing the resources involved in the system while meeting requirements of applications. From the acquisition of physical data to its transformation into valuable services or information, several steps must be performed, involving the various players in such a complex ecosystem. Support for decentralized data processing on IoT devices and other devices near the edge of the network, in combination with the benefits of cloud technologies has been identified as a promising approach to reduce communication overhead, thus reducing delay for time sensitive IoT applications. The interplay of IoT, edge and cloud to achieve the final goal of producing useful information and value-added services to end user gives rise to a management problem that needs to be wisely tackled. The goal of this work is to propose a novel resource management framework for edge-cloud systems that supports heterogeneity of both devices and application requirements. The framework aims to promote the efficient usage of the system resources while leveraging the Edge Computing features, to meet the low latency requirements of emergent IoT applications. The proposed framework encompasses (i) a lightweight and data-centric virtualization model for edge devices, (ii) a set of components responsible for the resource management and the provisioning of services from the virtualized edge-cloud resources