Sensor Function Virtualization to Support Distributed Intelligence in the Internet of Things

Abstract

It is estimated that-by 2020-billion devices will be connected to the Internet. This number not only includes TVs, PCs, tablets and smartphones, but also billions of embedded sensors that will make up the ''Internet of Things'' and enable a whole new range of intelligent services in domains such as manufacturing, health, smart homes, logistics, etc. To some extent, intelligence such as data processing or access control can be placed on the devices themselves. Alternatively, functionalities can be outsourced to the cloud. In reality, there is no single solution that fits all needs. Cooperation between devices, intermediate infrastructures (local networks, access networks, global networks) and/or cloud systems is needed in order to optimally support IoT communication and IoT applications. Through distributed intelligence the right communication and processing functionality will be available at the right place. The first part of this paper motivates the need for such distributed intelligence based on shortcomings in typical IoT systems. The second part focuses on the concept of sensor function virtualization, a potential enabler for distributed intelligence, and presents solutions on how to realize it.

Network Function Virtualization: State-of-the-Art and Research Challenges

“…NFV has been proposed as a key enabler of the IoT [192], [193]. The idea in [192] is to limit the functionality embedded in deployed sensors, and provide virtualized functions such as security, intelligence, computation and storage to the devices. These would take advantage of the scalable distribution capabilities of NFV as well as the configuration flexibility of SDN.…”

5G roadmap: 10 key enabling technologies

“…Specifically, as the IoT paradigm and its widespread applications create an explosive number of devices connected to the network, NFV would go a long way to reduce the cost of IoT devices by reducing the functionalities in devices and virtualizing those functions. For example, in [43], a sensor function virtualization that enables distributed intelligence in IoT is proposed through the deployment of modular blocks. More details of interaction between IoT and NFV will be discussed in Section 9.…”

Algorithms for advance bandwidth reservation in media production networks

“…NFV has been proposed as a key enabler of the IoT [195], [196]. The idea in [195] is to limit the functionality embedded in deployed sensors, and provide virtualized functions such as security, intelligence, computation and storage to the devices.…”

“…The idea in [195] is to limit the functionality embedded in deployed sensors, and provide virtualized functions such as security, intelligence, computation and storage to the devices. These would take advantage of the scalable distribution capabilities of NFV as well as the configuration flexibility of SDN.…”

“…IoT [195], [196] Architecture combining NFV and IoT, and an application scenario involving a virtualized sensor function ICN [198] exploits ICN to provide flexibility and dynamism in placing VNFs ETSI Use Cases [62], [73], [94], [95], [96], [97], [98], [100], [101], [102], [103], [104], [105], [106], [107], [108], [109], [110] Implementation, demonstrations and proofs of concepts based on the ETSI use cases More studies on the applicability of hardware acceleration to some NFs, and on the resulting trade-off between performance and flexibility…”