Cloud computing with its three key facets (i.e., IaaS, PaaS, and SaaS) and its inherent advantages (e.g., elasticity and scalability) still faces several challenges. The distance between the cloud and the end devices might be an issue for latencysensitive applications such as disaster management and content delivery applications. Service Level Agreements (SLAs) may also impose processing at locations where the cloud provider does not have data centers. Fog computing is a novel paradigm to address such issues. It enables provisioning resources and services outside the cloud, at the edge of the network, closer to end devices or eventually, at locations stipulated by SLAs. Fog computing is not a substitute for cloud computing but a powerful complement. It enables processing at the edge while still offering the possibility to interact with the cloud. This article presents a comprehensive survey on fog computing. It critically reviews the state of the art in the light of a concise set of evaluation criteria. We cover both the architectures and the algorithms that make fog systems. Challenges and research directions are also introduced. In addition, the lessons learned are reviewed and the prospects are discussed in terms of the key role fog is likely to play in emerging technologies such as Tactile Internet.

No abstract

Abstract-Wireless Sensor Networks (WSNs) are the key components of the emerging Internet-of-Things (IoT) paradigm. They are now ubiquitous and used in a plurality of application domains. WSNs are still domain specific and usually deployed to support a specific application. However, as WSNs' nodes are becoming more and more powerful, it is getting more and more pertinent to research how multiple applications could share a very same WSN infrastructure. Virtualization is a technology that can potentially enable this sharing. This paper is a survey on WSN virtualization. It provides a comprehensive review of the state-of-the-art and an in-depth discussion of the research issues. We introduce the basics of WSN virtualization and motivate its pertinence with carefully selected scenarios. Existing works are presented in detail and critically evaluated using a set of requirements derived from the scenarios. The pertinent research projects are also reviewed. Several research issues are also discussed with hints on how they could be tackled. perform computations and communicate [2]. The most obvious drawback of the current WSNs is that they are domain-specific and task-oriented, tailored for particular applications with little or no possibility of reusing them for newer applications. This strategy is inefficient and leads to redundant deployments when new applications are contemplated. With the introduction of the IoT, it is not unrealistic to envision that future WSN deployments will have to support multiple applications simultaneously. Index Terms-Virtualization is a well-established concept that allows the abstraction of actual physical computing resources into logical units, enabling their efficient usage by multiple independent users [3]. It is a promising technique that can allow the efficient utilization of WSN deployments, as multiple applications will be able to co-exist on the same virtualized WSN. Virtualization is a key technique for the realization of the Future Internet [4] and it is indeed quite pertinent to explore it in the context of WSNs.Virtualizing WSNs brings with it many benefits; for example, even applications that were not envisioned a priori may be able to utilize existing WSN deployments. A second, related benefit is the elimination of tight coupling between WSN services/applications and WSN deployments. This allows experienced as well as novice application developers to develop innovative WSN applications without needing to know the technical details of the WSNs involved. Another benefit is that WSN applications and services can utilize as well as be utilized by third-party applications. It can also help to define a business model, with roles such as physical WSN provider, virtual WSN provider and WSN service provider.The WSN virtualization concept can be applied to several interesting application areas. Recent advances in smart phones and autonomous vehicles [5] have made it possible to have multiple on-board sensors on them. Mobile crowd sensing is one area that can take advantage of virtuali...

No abstract

n the last few years, wireless sensor networks (WSNs) have become ubiquitous and are being used in a broad array of application domains, including healthcare, agriculture, surveillance, and security. These WSNs are composed of small-scale nodes that have the ability to sense, compute, and communicate [1]. While early sensor nodes were resource-constrained with limited capabilities, recent advances in sensor hardware technology have made it possible to produce sensor nodes that have more processing power and memory, and prolonged battery life.Virtualization is a key technique for the realization of the future Internet, and it is indeed quite pertinent to explore it in the context of WSNs. Virtualization makes it possible to present physical computing resources by abstracting them into logical units, enabling their efficient usage by multiple independent users, including multiple concurrent applications [2]. Furthermore, it allows for the deployment of applications that were not even envisioned during an infrastructure's initial deployment.To date, realizations of WSNs have been domain-specific and task-oriented. Applications are bundled with a WSN at the time of deployment, and it is next to impossible to use the same WSN for another application. This leads to redundant deployments and underutilization of these resources. There are two approaches to allow multiple applications to access deployed WSN resources. One is to allow multiple applications to share the data gathered from a WSN. In this approach, a sink/gateway node collects all the data from the WSN and shares it among multiple users. For example, in [3], WSNs are merged into the cloud by sending observed sensor data through a host manager that lies outside the WSN. The host manager simply collects the sensor data, profiles/aggregates it, and then allows multiple applications to use it for their own purposes.The second approach is to use the capabilities of the individual sensor nodes to execute multiple application tasks concurrently, and allow applications to group these sensor nodes together according to their requirements. The key difference between the two approaches is that the former approach allows the sharing of WSN data among multiple applications, while the latter allows sharing of WSN nodes by multiple applications. This article is focused on the second approach because it makes it possible to provision more innovative applications over the deployed WSNs, even applications that were not envisioned a priori. This will greatly improve the efficiency of deployed WSNs and will also encourage new business models.This article introduces the WSN virtualization concept, critically reviews the state of the art in WSN virtualization, and proposes a new early architecture that focuses on fixed WSNs. We illustrate the potential of the architecture by instantiating it for a fire monitoring scenario [4] in which multiple applications share the same WSN. We have built a prototype to demonstrate its feasibility and to measure its performance. We also identify ...