Search

The massive upsurge in production and demand of crowdsourced live streaming necessitates feasible and cost effective solutions in future cellular networks. Video multicast is one of the viable options to deliver live video using standardized multicast technologies, such as evolved Multimedia Broadcast Multicast Services (eMBMS) architecture. However, eMBMS being a service-oriented architecture, it does not support crowdsourced live video streaming from popular providers like Twitch, YouTube Live, and Facebook Live, where millions of viewers watch live streams. This article explores the possibilities for the crowdsourced service-less video multicast in fifth generation (5G) Radio Access Network (RAN). We highlight major limitations in the standard eMBMS architecture and propose a viable solution using Network Function Virtualization (NFV) and Mobile Edge Computing (MEC). The main challenges related to the initiation of an eMBMS service are presented, followed by our proposed network architecture, which augments the eMBMS architecture to initiate a service-less multicast video delivery for crowdsourced live videos. We evaluate and validate the proposed architecture using a real-world live video dataset. Our results depict the feasibility of the proposed architecture to multicast crowdsourced live streams and significant savings in cost and network consumption. [ABSTRACT FROM AUTHOR]

Benefitting from large-scale training datasets and the complex training network, Convolutional Neural Networks (CNNs) are widely applied in various fields with high accuracy. However, the training process of CNNs is very time-consuming, where large amounts of training samples and iterative operations are required to obtain high-quality weight parameters. In this paper, we focus on the time-consuming training process of large-scale CNNs and propose a Bi-layered Parallel Training (BPT-CNN) architecture in distributed computing environments. BPT-CNN consists of two main components: (a) an outer-layer parallel training for multiple CNN subnetworks on separate data subsets, and (b) an inner-layer parallel training for each subnetwork. In the outer-layer parallelism, we address critical issues of distributed and parallel computing, including data communication, synchronization, and workload balance. A heterogeneous-aware Incremental Data Partitioning and Allocation (IDPA) strategy is proposed, where large-scale training datasets are partitioned and allocated to the computing nodes in batches according to their computing power. To minimize the synchronization waiting during the global weight update process, an Asynchronous Global Weight Update (AGWU) strategy is proposed. In the inner-layer parallelism, we further accelerate the training process for each CNN subnetwork on each computer, where computation steps of convolutional layer and the local weight training are parallelized based on task-parallelism. We introduce task decomposition and scheduling strategies with the objectives of thread-level load balancing and minimum waiting time for critical paths. Extensive experimental results indicate that the proposed BPT-CNN effectively improves the training performance of CNNs while maintaining the accuracy. [ABSTRACT FROM AUTHOR]

Cloud computing has emerged as the leading paradigm for information technology businesses. Cloud computing provides a platform to manage and deliver computing services around the world over the Internet. Cloud services have helped businesses utilize computing services on demand with no upfront investments. The cloud computing paradigm has sustained its growth, which has led to increase in size and number of data centers. Data centers with thousands of computing devices are deployed as back end to provide cloud services. Computing devices are deployed redundantly in data centers to ensure 24/7 availability. However, many studies have pointed out that data centers consume large amount of electricity, thus calling for energy-efficiency measures. In this survey, we discuss research issues related to conflicting requirements of maximizing quality of services (QoSs) (availability, reliability, etc.) delivered by the cloud services while minimizing energy consumption of the data center resources. In this paper, we present the concept of inception of data center energy-efficiency controller that can consolidate data center resources with minimal effect on QoS requirements. We discuss software- and hardware-based techniques and architectures for data center resources such as server, memory, and network devices that can be manipulated by the data center controller to achieve energy efficiency. [ABSTRACT FROM PUBLISHER]

Data Center Networks (DCNs) constitute the communication backbone for the cloud computing paradigm. Recently, network connectivity analysis in terms of reliability has received attention from the network research community. The traditional network features are useful; however, they are insufficient to determine how well-connected or well-designed a DCN is against the node or link removals. In this letter, we present a connectivity analysis of three well-known DCN architectures, namely: (a) ThreeTier, (b) FatTree, and (c) DCell. Our analysis reveals that the classic connectivity measures are inadequate for evaluating DCN connectivity. Therefore, we propose μ-A2TR, a novel metric to characterize network connectivity in the case of node or link failures. Experimental results reveal that the DCNs exhibit a moderate level of connectivity in the case of random node removals. However, connectivity decays abruptly when considering the targeted nodes removal. Moreover, the connectivity analysis depicts significant differences among the considered DCNs. [ABSTRACT FROM PUBLISHER]

SUMMARY Data centers are experiencing a remarkable growth in the number of interconnected servers. Being one of the foremost data center design concerns, network infrastructure plays a pivotal role in the initial capital investment and ascertaining the performance parameters for the data center. Legacy data center network (DCN) infrastructure lacks the inherent capability to meet the data centers growth trend and aggregate bandwidth demands. Deployment of even the highest-end enterprise network equipment only delivers around 50% of the aggregate bandwidth at the edge of network. The vital challenges faced by the legacy DCN architecture trigger the need for new DCN architectures, to accommodate the growing demands of the 'cloud computing' paradigm. We have implemented and simulated the state of the art DCN models in this paper, namely: (a) legacy DCN architecture, (b) switch-based, and (c) hybrid models, and compared their effectiveness by monitoring the network: (a) throughput and (b) average packet delay. The presented analysis may be perceived as a background benchmarking study for the further research on the simulation and implementation of the DCN-customized topologies and customized addressing protocols in the large-scale data centers. We have performed extensive simulations under various network traffic patterns to ascertain the strengths and inadequacies of the different DCN architectures. Moreover, we provide a firm foundation for further research and enhancement in DCN architectures. Copyright © 2012 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]