The recent availability of smart network interface cards (smart NICs) and Data Processing units (DPUs) providing hardware-accelerated networking and computing functionalities is opening the way towards new application...
详细信息
ISBN:
(数字)9781665497268
ISBN:
(纸本)9781665497268
The recent availability of smart network interface cards (smart NICs) and Data Processing units (DPUs) providing hardware-accelerated networking and computing functionalities is opening the way towards new applications and use cases beyond the traditional data center scenarios. In this paper, three different use cases for edge scenarios that leverage on the innovative programmability enabled by DPUs are presented and discussed. The first use case focuses on a pervasive monitoring infrastructure to support accurate and decentralized network awareness for low-latency 5G services The second one focuses on the implementation of power-efficient edge-to-cloud continuum. The third use case refers to effective network security functions at the DPU.
Internet usability is expanded form just human-to-human interactions toward different communication types, while the communication itself is shifting from the host-centric model to the content-centric paradigm. The 5G...
详细信息
Internet usability is expanded form just human-to-human interactions toward different communication types, while the communication itself is shifting from the host-centric model to the content-centric paradigm. The 5G and beyond networks promise not only to support such changes but also to provide massive data exchange and connectivity with high reliability. The next-generation networking technologies are the key enabled for 5G that aim at building a new ecosystem. One promising piece of this ecosystem is the information-centric network (ICN), which is a future network architecture that tends to tackle the current host-centric model issues. It natively supports several features, including abstraction content naming and transparent in-network content caching that contribute to improve network performance, reduce traffic, and improve the latency. In this article, we first provide a potential road map by introducing different next-generation active technologies to enable the big picture of 5G, including mobile-edge computing (MEC), software-defined networking (SDN), and network function virtualization (NFV). Then, we discuss the need for ICN and its coexistence within this ecosystem. Later, we present an in-depth review of the recent content naming schemes and a comprehensive review of in-network content caching solutions. We classify these solutions into different classes based on the used technologies and their working principle. Finally, we highlight some research challenges and propose promising directions for the research community.
The Internet of Things (IoT) is key enabling technology for future Internet, and aims at connecting heterogeneous devices and objects using wireless technologies. This connectivity results in creating a massive number...
详细信息
The Internet of Things (IoT) is key enabling technology for future Internet, and aims at connecting heterogeneous devices and objects using wireless technologies. This connectivity results in creating a massive number of heterogeneous devices, and a large amount of content, thereby creating new challenges regarding content, service, and device naming. Current Internet applications and use cases are now shifting toward the content-centric paradigm, where the content is the key element in the infrastructure. In this article, we design a hybrid multilayer naming scheme with multi-component hierarchical and attribute-value components for a data-centric Internet of Things. The proposed scheme targets smart IoT applications and provides built-in scalability, efficient routing, and security features. We incorporate a variable-length encoding method, with a prefix-labeling scheme in order to describe hierarchical location names with various embedded semantic functionalities. Moreover, the scheme supports fast local IoT communication using Name-to-Code translation concept, as well as multi-source content retrieval through in-network function. To evaluate the performance and prove the efficiency of the proposed scheme, we carry extensive experiments, using ndnSIM network simulator. The analysis shows that our proposed hybrid naming is inherently better than the existing stateof-the-art solutions, and drastically reduces the memory consumption, lookup time, routing and forwarding overhead, and enhances the overall loT communication.
The energy resources of Machine-to-Machine (M2M) devices need to last as much as possible. Data aggregation is a suitable solution to prolong the network lifetime, since it allows the devices to reduce the amount of d...
详细信息
ISBN:
(纸本)9781479966066
The energy resources of Machine-to-Machine (M2M) devices need to last as much as possible. Data aggregation is a suitable solution to prolong the network lifetime, since it allows the devices to reduce the amount of data traffic. In M2M systems, the M2M platform and the Constrained Application Protocol (CoAP) enable multiple entities to send concurrent data-requests to the same capillary network. For example, in a Smart Metering scenario, there are devices measuring the electricity consumption of an entire building. The supplier company requests all devices to send the data updates every 1800 seconds (i.e., 30 minutes). On the other hand, a resident requests his/her devices to communicate every 600 seconds (i.e., 10 minutes). These concurrent data-requests create heterogeneous groups over the same capillary network, since each group might be able to execute different in-network functions and to have a unique temporal-frequency of communication. However, the traditional data aggregation solutions designed for periodic monitoring assume the execution of a single static data-request during all network lifetime. This makes the traditional data aggregation solutions not suitable for M2M environments. To fill this gap, this paper presents Data Aggregation for Multiple Groups (DAMiG), which is designed to provide Data Aggregation for heterogeneous and concurrent sets of CoAP data-requests. DAMiG explores the group communication periodicity to perform internal and external-group traffic aggregation. To achieve that, DAMiG computes a suitable aggregation structure and applies statistical and merger aggregation functions along the path. DAMiG is able to reduce the energy consumption in scenarios with single or several concurrent CoAP data-requests. Moreover, the selection of internal and external-group paths takes into account the residual energy of the nodes, avoiding the paths with low residual energy.
暂无评论