Computation Offloading contributes towards moving to a Mobile Cloud Computing paradigm. In this work, a two-level resource allocation and admission control mechanism for a cluster of edge servers, offers an alternative choice to mobile users for executing their tasks. At the lower level, the behavior of edge servers is modeled by a set of linear systems and linera controllers are designed to meet the system's constraints and QoS metrics, while at the upper level, an optimizer tackles the problems of load balancing and application placement towards maximizing the number the offloaded requests.
Scientific workflows enable to conduct analysis on large datasets and perform complex scientific simulations. These workflows are often mapped onto distributed computational infrastructures to speed up their execution. Prior execution, a workflow structure may suffer transformations to accommodate the computing infrastructures. However, these transformations may cause workflow imbalance because of runtime or data imbalance. To mitigate these imbalances, in this paper we propose an autonomic data-throttling approach to compute how data transmission must be throttled throughout workflow jobs. Our approach relies on structural analysis of Petri nets, obtained by model transformation of data-intensive workflows, and Linear Programming techniques.
In this paper, we aim to understand the dynamics of bundling sale strategy and under what situations it will be more attractive than the separate sales. We focus on online service markets that exhibit network effect. We provide mathematical models to capture the interactions between buyers and sellers, analyze the market equilibrium and its stability, and formulate an optimization framework to determine the optimal sale strategy for the service provider. We analyze the impact of the key factors such as network effects, operating costs, as well as the variance and correlation of customers' valuations towards these services.
The diffusion of personal smart devices capable of capturing and analyzing user's data has encouraged the growth of novel sensing methodologies. In this context, simple Human Activity Recognition (HAR) techniques can be directly implemented on mobile devices; however, when complex activities need to be analysed timely, users' personal devices can operate as part of a more complex architecture. We propose a multi-device HAR framework that exploits the fog computing paradigm to move heavy computation from the sensing layer (wrist-worn devices), to intermediate devices (smartphones), and then to the cloud. Results show the effectiveness of the framework on a real-world scenario.
Mobile crowd sensing (MCS) has been utilized to develop several context aware applications that obtain knowledge about the environment from large set of devices. We present case of MCS to be potentially used for much more demanding applications such as real time perpetrator tracking by online mining of images taken from nearby mobile surrogates. ContextAiDe architecture, a combination of API, middleware, and optimization engine, that enables optimization of MCS operational overheads in addition to computation and communication requirements of the MCS application to enable real time operation under demanding workloads. Results of sample run of perpetrator tracking app are presented.
Standard security protocols are characterized by high computational complexity that is unsuitable to networks of low-power devices. The typical solution based on cloud services that facilitate deployment, intermediate all messages among things and enable secure communications has the disadvantage of requiring permanent Internet connectivity even for things connected over a local network. This paradigm is inappropriate in several scenarios, hence we propose an efficient fog-based system that enables secure communications and preserves easy management of cloud-assisted IoT. The proposal is based on an original lightweight proxy re-encryption scheme that can be executed even by large networks of low-power devices.
In this paper, we propose an automatic labeling approach to distinguish crisis-relevant tweets while differentiating source types (e.g., government or personal accounts) simultaneously. We first analyze and identify tweet-specific linguistic, sentimental and emotional features based on statistical topic modeling. Then,we design a novel correlative convolutional neural network which uses a shared hidden layer to learn effective representations of the multi-faceted features. The model can discover salient information robust to the variations and noises in tweets and sources. To obtain a bird-view of crisis event, we further develop an approach to automatically summarize key information of identified tweets.
In this paper, we focus on computation offloading over multi-cloudlets environment. We consider several mobile users with different energy and latency constrained tasks that can be offloaded over cloudlets. We investigate offloading policy that decides which tasks should be offloaded and selects the assigned cloudlet, accordingly with network and system resources. The objective is to minimize the execution time and the energy consumption. We propose a distributed relaxation heuristic based on Lagrangian decomposition. Numerical results show that our policy achieves a good offloading solution quickly, and can achieve better performances for large-scale scenarios compared to alternatives approaches from the literature.
This article presents a comprehensive survey on Fog Computing for the IoT. Hereinafter, we detail the principles characterizing Fog Computing and provide the historical background of this paradigm. We also highlight the IoT application domains that may benefit from it and review the existing literature for each of them. Furthermore, we analyze the challenges afflicting the Fog, report how the research community is facing them, and point out the open issues and future research directions. Last but not least, we provide an overview of the main Fog Computing platforms for the Internet of Things.
Search engines cannot answer time and location-specific queries; this often requires humans on-site. While community question answering (CQA) platforms are popular, few exceptions consider users physical locations. Here, we present CloseUp, our prototype for the integration of community-driven live search into a Google-like search experience. We bridge the gap between Web search and CQA, namely the formulation of search requests and the expected response times. CloseUp features a deep learning pipeline to translate relevant queries into questions. CloseUp provides a mobile application for submitting and replying to questions. Using a field study, we evaluated the feasibility of our approach.
Despite the growing body of research focused on understanding Knowledge Intensive Processes (KIPs), the research question on how to measure the performance of KIPs and of the knowledge workers involved is still open. In this paper, we address it with a proposal to enable performance management of KIPs: An ontology that allows us to define process performance indicators in the context KIPs, and a methodology that builds on the ontology and the concepts of lead and lag indicators. Both the ontology and the methodology have been applied to a case study of a real ICT Outsourcing Company in Brazil.
Business process improvement is challenging many organizations. As long as there is a process, it must be improved. Nowadays, improvement initiatives are driven by professionals. This is no longer practical because people cannot fathom the abundance of data generated. Here, we propose improving processes using ubiquitous computing and business decisions. We introduce ubiquitous decisions-aware business processes that pervade the physical space, analyze the ever-changing environments and make decisions accordingly. We explain how they can be built and used for improvement. Our approach can be a valuable improvement option to help participants focus on the crucial rather than the menial tasks.
Detecting concurrency relations between events is a fundamental primitive underpinning a range of process mining techniques. Existing approaches identify concurrency relations at the level of event types under a global interpretation. If two event types are declared to be concurrent, every occurrence of one event type is deemed to be concurrent to one occurrence of the other. In practice, this interpretation is too coarse-grained and leads to over-generalization. This paper proposes a finer-grained approach, whereby two event types may be deemed to be in a concurrency relation relative to one state of the process, but not relative to other states.