Edge computing processes and stores data on user devices located at the edge of the Internet to support latency-sensitive services such as autonomous vehicles, drones, and so on. This adopts container technology to qu...
详细信息
ISBN:
(纸本)9781665423830
Edge computing processes and stores data on user devices located at the edge of the Internet to support latency-sensitive services such as autonomous vehicles, drones, and so on. This adopts container technology to quickly initialize and deploy applications. Containers interact with each other through Plugins implemented based on Container Network Interface (CNI). In particular, the network latency is different depending on the plugin. In this paper, we study plugins to provide low-latency edge services and evaluate the effect on data processing performance when multiple plugins are used at the same time. The SR-IOV plugins showed twice throughput than the kernel-based virtual network interface. Also, it had more than 50% bandwidth regardless of the addition of other virtual network interfaces on a single connection.
For flexibly providing bandwidth-intensive services to a number of multi-radio users (MR-UEs) in cellular heterogeneous networks, Software Defined Networking (SDN) emerges as an evolutionary notion because of its cent...
详细信息
ISBN:
(纸本)9781538650417
For flexibly providing bandwidth-intensive services to a number of multi-radio users (MR-UEs) in cellular heterogeneous networks, Software Defined Networking (SDN) emerges as an evolutionary notion because of its centralized management. However, such SDN based fully-centralized control architecture of cellular networks still suffer from scalability issues, fault tolerance, latency problems etc. Besides, fog computing is introduced as an extension of cloud computing from core to the edges of network, bringing real-time and latency-sensitive services close to the users. Furthermore, a number of multicast users can be served through Device-to-device (D2D) communication without effecting the backhaul network. In this paper, we design a novel semi-centralized architecture for eMBMS/WLAN heterogeneous network which is assisted by multi-level control plane and fog nodes, to efficiently serve D2D multicast users. Given the access points have capability to be transformed into configurable fog nodes, we believe that popular multicast content can be cached by fog nodes to serve MR-UEs in less time. Under the supervision of one controller, multiple local controllers collaborate towards reliable multicasting in SDN-based heterogeneous networks.
Recently, edge computing has received considerable attention as a promising solution to provide deep learning-based video analysis services in real-time. However, due to the limited computation capability of the data ...
详细信息
ISBN:
(纸本)9781728167589
Recently, edge computing has received considerable attention as a promising solution to provide deep learning-based video analysis services in real-time. However, due to the limited computation capability of the data processing units (such as CPUs, GPUs, and specialized accelerators) embedded in the edge devices, the question of how to use the limited resources of the edge devices is one of the most pressing issues affecting deep learning-based video analysis service efficiency. In this paper, we introduce a practical approach to optimize deep learning object detection at the edge devices embedding CPUs and GPUs. The proposed approach adopts TVM, an automated end-to-end deep learning compiler that automatically optimizes deep learning workloads with respect to hardware-specific characteristics. In addition, task-level pipeline parallelism is applied to maximize resource utilization of the CPUs and GPUs so as to improve overall object detection performance. Through experiment results, we show that the proposed approach achieves performance improvement for detecting objects on multiple video streams in terms of frame per second.
With the advent of disruptive programmable network technologies such as SDN and NFV, network evolution has been spurred by accommodating cloud service infrastructure inside the network. Accordingly the key issues for ...
详细信息
ISBN:
(纸本)9781479967865
With the advent of disruptive programmable network technologies such as SDN and NFV, network evolution has been spurred by accommodating cloud service infrastructure inside the network. Accordingly the key issues for the service orchestration and harmonization to provide scalable and reliable application services on that environment would be investigated through the PoC service scenarios.
Cloud computingtechnology enables service to add special characteristics such as remote execution, service mobility and virtual workspace in a handheld device. However, it has a limitation on application areas for th...
详细信息
ISBN:
(纸本)9788988678558
Cloud computingtechnology enables service to add special characteristics such as remote execution, service mobility and virtual workspace in a handheld device. However, it has a limitation on application areas for the reason that mobile applications distinctly need a help of the host computer. To eliminate those threats, hash chain based approaches are proposed [1][2]. However, it also has some restriction that a mobile device sends request messages to each server when the device needs a multi-entity co-operative work. Therefore, this paper introduces a concept of secure entrustment, which provides most near and trusted servers whenever your mobile devices want to use a trusted computing.
Since the late 1980s the world is working towards connectivity and convergence. In the last three decades, the convergence of information resources has happened. However to achieve a true convergence the information a...
详细信息
Modern computer game engine has evolved to become a complete suite of virtual world constructing tools and runtime environment. The latter is usually a tightly integrated framework of 3D rendering engine, scripting en...
详细信息
ISBN:
(纸本)9780769534077
Modern computer game engine has evolved to become a complete suite of virtual world constructing tools and runtime environment. The latter is usually a tightly integrated framework of 3D rendering engine, scripting engine, physics simulation and networking. This framework design as well as individual component implementations decides the general type of games that could be composed by it. This article proposes a game engine framework called ParaEngine for developing games based on distributed game world data and logic. Its framework design is based on a modified version of Simulation Theory about the human brain. It has long been observed that our brain is both a distributed computing environment and a theatre of multimedia. The analogy of human cognition to simulation system has been applied to the proposed game engine to construct distributed Internet games. The implementation is illustrated in an RPG game demo called National Heroes.
Artificial Intelligence is growing very fast. Along with the development of artificial intelligence, Ambient intelligence is also developing, which can identify and transmit various activities such as people's dai...
详细信息
ISBN:
(纸本)9781728108933
Artificial Intelligence is growing very fast. Along with the development of artificial intelligence, Ambient intelligence is also developing, which can identify and transmit various activities such as people's daily life and work. Ambient Intelligence has developed into convergencetechnology of Artificial Intelligence, Cloud computing and Internet of Things. We will discuss each technology. In this paper, we propose Personalized Ambient Intelligence which focuses on personalization based on the necessary element technologies for Ambient Intelligence(AmI). The proposed Personalized Ambient Intelligence should consist of Ambient Intelligence environment, AmI Common Data Model, Context Classification, AmI Sensor Technologies and Ambient intelligence situation. Among them, Personalized Ambient Situation should be classified into 4 situation as Home, Business, Driving, Moving. Using Personalized Ambient Intelligence presented in this paper, personalized context awareness is likely to be possible, and prediction of the next context will also be possible. And Finally, we will look at a brief summary of our definition of Personalized Ambient Intelligence and the areas that can be used in the future and conclude with future research.
The particle swarm optimization (PSO) is a population-based strategy for global optimization. A rapid decrease of diversity in the iterative procedure leads the PSO to suffer from premature convergence on multimodal p...
详细信息
ISBN:
(纸本)9781467308946;9788994364261
The particle swarm optimization (PSO) is a population-based strategy for global optimization. A rapid decrease of diversity in the iterative procedure leads the PSO to suffer from premature convergence on multimodal problems. This paper presents a novel variant of the PSO (PSO-CAR). It uses an animal foraging strategy with attraction and repulsion. This strategy guarantees a high diversity of the swarm to protect the PSO from premature convergence. The basic PSO (bPSO), genetic algorithm (GA) and two other variants of the original PSO are employed here to evaluate the effectiveness of the PSO-CAR using 6 standard numerical functions. Simulation results reveal that the proposed PSO-CAR outperforms the other approaches in solving multimodal problems.
In this paper, gives the definition of complex fuzzy matrix, and study its convergence problems based on the fuzzy matrix theory, which included the convergence in norm and the convergence in Power, some important con...
详细信息
ISBN:
(纸本)9783319191058;9783319191041
In this paper, gives the definition of complex fuzzy matrix, and study its convergence problems based on the fuzzy matrix theory, which included the convergence in norm and the convergence in Power, some important conclusions are obtained, to build and to improve the solid foundation for complex fuzzy matrix theory.
暂无评论