Wirelessly streaming high quality 360 degree videos is still a challenging problem. When there are many users watching different 360 degree videos and competing for the computing and communication resources, the strea...
详细信息
ISBN:
(数字)9798350363999
ISBN:
(纸本)9798350364002
Wirelessly streaming high quality 360 degree videos is still a challenging problem. When there are many users watching different 360 degree videos and competing for the computing and communication resources, the streaming algorithm at hand should maximize the average quality of experience (QoE) while guaranteeing a minimum rate for each user. In this paper, we propose a cross layer optimization approach that maximizes the available rate to each user and efficiently uses it to maximize users' QoE. Particularly, we consider a tile based 360 degree video streaming, and we optimize a QoE metric that balances the tradeoff between maximizing each user's QoE and ensuring fairness among users. We show that the problem can be decoupled into two interrelated subproblems: (i) a physical layer subproblem whose objective is to find the download rate for each user, and (ii) an application layer subproblem whose objective is to use that rate to find a quality decision per tile such that the user's QoE is maximized. We prove that the physical layer subproblem can be solved optimally with low complexity and an actor-critic deep reinforcement learning (DRL) is proposed to leverage the parallel training of multiple independent agents and solve the application layer subproblem. Extensive experiments reveal the robustness of our scheme and demonstrate its significant performance improvement compared to several baseline algorithms.
Blockchains have seen a recent rise in popularity as a generic solution for trustless distributed applications across a wide range of industries. However, blockchain protocols have faced scalability issues in applicat...
详细信息
ISBN:
(纸本)9781728183701
Blockchains have seen a recent rise in popularity as a generic solution for trustless distributed applications across a wide range of industries. However, blockchain protocols have faced scalability issues in applications involving a growing number of participants. In this paper we instantiate and evaluate StakeCube, a proposal for a scalable shard-based distributed ledger. We further detail and tune a byzantine agreement algorithm suited for StakeCube's sharding structure, and we experimentally study and asses its performance, especially regarding scalability. We were successfully able to run StakeCube with up to 5000 participants, confirming up to 1100 bytes/s of transaction, with a confirmation time starting at 200 seconds. Finally, we use StakeCube in a large scale energy marketplace application, and show that a node running on a Raspberry Pi Zero is able to handle the load without issues.
The recent development in the field of power grid has observed that the affect of distributed energy resourcefulness's, electric vehicles, plug-in hybrid electric vehicles, and smart gadgets are favourable to the ...
详细信息
ISBN:
(数字)9781728185293
ISBN:
(纸本)9781728185309
The recent development in the field of power grid has observed that the affect of distributed energy resourcefulness's, electric vehicles, plug-in hybrid electric vehicles, and smart gadgets are favourable to the environment. It is economical. And also it is reliable. The needs can be enhanced by executing coordinated smart controls. Coordinated control's absence may have some negative effects, such as reduced lifelong service of power distribution components and in distribution transformers. This paper represents a smart houses with the application of energy management system & smart grid that permits organized control for a house resourcefulness's without consumer bother while minimizes overloading and overheating of the distributed substructure. Paper also aims an architectural structure and an operational model for house Energy Management System (EMS), a structure in a house that helps household inhabit in working house equipment to gain optimal energy utilization. The architectural system also looks a functional relation between parts and the whole system of the structure and brings into the notice with the smart grid which are the Distribution System Operators (DSOs) and Energy Service Providers (ESPs).
In recent years, classification with big data sets has become one of the latest research topic in machine learning. distributed classification have received much attention from industry and academia. Recently, the Alt...
详细信息
Nowadays, data parallelism has been widely applied to train large datasets on distributed deep learning clusters, but it has suffered from costly global parameter updates at batch barriers. Performance imbalance among...
详细信息
Nowadays, data parallelism has been widely applied to train large datasets on distributed deep learning clusters, but it has suffered from costly global parameter updates at batch barriers. Performance imbalance among worker instances, introduced by uneven workload partitioning or biased resource allocation, can cause straggly workers, which can lead to severe impacts on both training speed and result accuracy. This paper studies the issue focusing on the tradeoff between training speed and result accuracy. We propose Cooperate Grouping parallel (CGP), a hybrid parameter update solution that allows the flexibility of both synchronous and asynchronous update schemes. We introduce a novel Cooperate Worker Grouping Problem (CWGP) that seeks a task grouping configuration that leads to maximum model accuracy and holds customized training speed guarantees. We propose an evolution-based Pareto local searching algorithm to compute efficient worker grouping configurations. Comprehensive evaluation results are presented to demonstrate the effectiveness of CGP under both persistent and fluctuant imbalances. The proposed method alleviates the imbalance impacts without introducing extra adjustment over-heads.
With rapid increase in the number of Phasor Measurement Units (PMUs) in the electric grid, massive volumes of monitoring data are expected to overwhelm the data preprocessors at centralized computing facilities. This,...
详细信息
ISBN:
(纸本)9781728161273
With rapid increase in the number of Phasor Measurement Units (PMUs) in the electric grid, massive volumes of monitoring data are expected to overwhelm the data preprocessors at centralized computing facilities. This, along with the requirements of lower latency and increased resilience to data anomalies advocates for distributed architectures for data conditioning and processing. To that end, in this paper, we present a fog-computing-based hierarchical approach for distributed detection and correction of anomalies in PMU data. In our proposed approach, each fog node, responsible for real-time data preprocessing, is dynamically assigned a smaller group of PMU signals with similar modal observabilities using software-defined-networking (SDN). The SDN controller residing at a central node feeds on the modeshapes estimated from the signals recovered at each fog node, for running the PMU-grouping algorithm. Grouping ensures adequate denseness of each signal set and guarantees data recovery under corruption. Also, the grouping is soft-real-time, infrequent, and triggered only upon a change in operating condition and therefore, heavily relieves the computational burden off the central node. The effectiveness of the proposed approach is demonstrated using simulated data from the IEEE 5-area 16-machine test system.
Algorithms for temporal property graphs may be time-dependent (TD), navigating the structure and time concurrently, or time-independent (TI), operating separately on different snapshots. Currently, there is no unified...
详细信息
ISBN:
(纸本)9781728129037
Algorithms for temporal property graphs may be time-dependent (TD), navigating the structure and time concurrently, or time-independent (TI), operating separately on different snapshots. Currently, there is no unified and scalable programming abstraction to design TI and TD algorithms over large temporal graphs. We propose an interval-centric computing model (ICM) for distributed and iterative processing of temporal graphs, where a vertex's time-interval is a unit of data-parallel computation. It introduces a unique time-warp operator for temporal partitioning and grouping of messages that hides the complexity of designing temporal algorithms, while avoiding redundancy in user logic calls and messages sent. GRAPHITE is our implementation of ICM over Apache Giraph, and we use it to design 12 TI and TD algorithms from literature. We rigorously evaluate its performance for diverse real-world temporal graphs as large as 131M vertices and 5.5B edges, and as long as 219 snapshots. Our comparison with 4 baseline platforms on a 10-node commodity cluster shows that ICM shares compute and messaging across intervals to out-perform them by up to 25x, and matches them even in worst-case scenarios. GRAPHITE also exhibits weak-scaling with near-perfect efficiency.
The proceedings contain 26 papers. The topics discussed include: deep learning detection method of encrypted malicious traffic for power grid;evaluation of voltage sag severity in provincial power grid;distributed mul...
ISBN:
(纸本)9780738105000
The proceedings contain 26 papers. The topics discussed include: deep learning detection method of encrypted malicious traffic for power grid;evaluation of voltage sag severity in provincial power grid;distributed multi-factor electricity transaction match mechanism based on blockchain;an overview: data security mechanism of power terminal in edge computing;a blockchain-based distributed controllable electricity transaction match system;recommendation and election expert system for rotating machinery fault diagnosis based on the combination of rules and examples;energy-use internet and friendly interaction with power grid: a perspective;a case study of developing an intelligent management system for energy internet;and modeling of short-term memory effect in electric double layer capacitor with graphene-based electrode.
The IEEE PerDL workshop aims to address all relevant technologies, researches and discoveries in the field of distributed and pervasive Deep Learning (DL) techniques. The usage of deep learning techniques is becoming ...
The IEEE PerDL workshop aims to address all relevant technologies, researches and discoveries in the field of distributed and pervasive Deep Learning (DL) techniques. The usage of deep learning techniques is becoming more and more widespread and needs to be properly incorporated into pervasive and distributedcomputing and communication architectures, in order to foster a fully connected environment of intelligent systems. The great development of Internet of Things devices and applications is fostering more and more pervasive communication paradigms, but smart devices and gateways need further capabilities to become fully connected intelligent things. What is still lacking is an in-depth study of the modalities and technologies to properly integrate all deep learning aspects and algorithms in all possible distributed and pervasive systems, making them actual intelligent distributed systems. The main aim is to support fast and on-the-field DL-based computations that may foster novel and fast services in the near future contexts, such as smart cities, smart agriculture, smart health, smart distance education, smart automotive, industry 4.0, and many more. On the other hand, DL could help in finding out better distributed and parallel configurations to increase the performance of pervasive computing itself. The objective of the PerDL workshop is to encourage the integration between the distributed and pervasive computing community and the deep artificial neural network community. This with the aim to foster the development of more and more widespread deep learning techniques in all computing and communication systems, as well as to increase pervasive computing performance by means of deep learning strategies.
With the rapid development of edge computing technology, the application of edge computing in smart grids has become more and more extensive. But edge computing has not yet been applied to the operation control of dis...
详细信息
With the rapid development of edge computing technology, the application of edge computing in smart grids has become more and more extensive. But edge computing has not yet been applied to the operation control of distributed power generation microgrid systems. This article proposes a microgrid-oriented edge computing architecture. First, we introduce the main functions of edge-cloud collaboration. Then we explain the construction plan of the architecture, including the realization of data processing, network communication and security mechanisms. Finally, we introduce the architecture application practice in a rural community in Central China.
暂无评论