Wireless sensor network virtualization: early architecture and research perspectives

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 ...

Virtualization in Wireless Sensor Networks: Fault Tolerant Embedding for Internet of Things

“…Some of the initial works on the virtualization focuses on service-oriented network architecture to optimize sensing resources. An fourlayered architecture for virtualization in WSNs has been suggested, based on reducing redundant deployment of sensor networks for different IoT applications [8]. Another fivelayered virtualization architecture has been explored to support network diversity and increase resource utilization in IoT [2].…”

“…Some use cases have conceptualized the virtual networking in IoT. An architecture has been suggested for virtualization of WSNs, to reduce redundant deployment of sensor networks for different IoT applications [8]. The architecture includes four layers considering physical, virtual sensor, virtual sensor access, and application.…”

Energy-aware dynamic resource allocation in virtual sensor networks

“…In this context, virtualization is a promising technique to achieve an efficient reuse of general purpose wireless sensor networks to dynamically support multiple applications and services [1]. The key idea behind this approach, which often goes under the names of Virtual Sensor Networks (VSN), is to abstract away "physical resources" including node processing/storage capabilities, available communication bandwidth and routing protocols, which can then be "composed" at a logical level to support usage by multiple independent users and even by multiple concurrent applications.…”

“…The key idea behind this approach, which often goes under the names of Virtual Sensor Networks (VSN), is to abstract away "physical resources" including node processing/storage capabilities, available communication bandwidth and routing protocols, which can then be "composed" at a logical level to support usage by multiple independent users and even by multiple concurrent applications. This new paradigm has stimulated research efforts in the field of novel programming abstractions at the node level and management framework at the network level to support multiple applications over a shared physical infrastructure [1], [2].…”

Alternative Resource Allocation Model for Dynamic Resource Sharing WSN Systems

“…In general, different types of WSN virtualization methods exist, node-based virtualization (NodBV), network-based virtualization (NetBV) [1][2][3]. But the methods are not suitable for IoT systems that meet the demands of a large number of clients.…”