云媒体在视频会议中应用的关键技术研究

云媒体服务方式是通信网络中替代传统媒体服务器的发展方向,其主要实现语音、视频以及相关业务应用的媒体处理功能.通过对视频会议产品发展需求的整理,从现有云媒体整体架构出发进行了分析,对通过云媒体的方式实现视频会议产品的关键技术进行了研究.     

一种云媒体服务器的系统设计及应用分析

提出了一种基于云计算的媒体服务器系统架构,提供了一种云媒体服务器的调度方法,并对云媒体的应用前景进行了分析。基于本文所述的云媒体服务器系统,能根据当下的业务流量提供自适应性的性能管理能力,使得由之构建的媒体服务能力池中设备资源利用率实现最大化,从而降低运营商业务网络的运营成本。     

基于IP网络的媒体点播系统设计

设计了基于IP网络的媒体点播系统,对媒体服务器的主要性能参数进行了测试,分析了在特定环境下影响服务器性能的主要原因,为优化媒体服务器提供了依据.     

下一代网络中媒体服务器的设计与实现

媒体服务器是下一代网络中的重要设备,负责提供专用的媒体资源。介绍了一种支持多协议、可扩展、高性能的媒体服务器的设计与实现方法。该媒体服务器支持SIP和H.248两种控制协议,采用多线程技术提高系统的性能。     

IP媒体服务器关键功能与实现机制

IP Media Server(IP媒体服务器)是下一代网络解决方案的一个重要组成部分,它位于网络的业务层,为各类业务提供所需的媒体资源和媒体服务。文章提出IP Media Server需要实现的关键功能,并对实现的关键技术进行分析,给出音频编码支持、回声消除、时延和抖动、放音机制、智能混音处理等技术的实现机制,有效地解决了媒体服务器开发中的技术难点。     

因特网上的媒体流传输——媒体服务器

本介绍了媒体流传输系统的组成、客户工作站、媒体服务器以及媒体服务器的网络连接。     

AcroMedia媒体服务器的设计与实现

媒体服务器是下一代网络平台中提供多媒体资源的关键设备。提出一种基于DSP的可扩展、高性能的AcroMedia媒体服务器的设计实现方法。阐述了Acro Media的系统架构,并详细描述了硬件关键技术和MCU的工作原理,阐述了软件模块划分以及DSP资源的控制流程。根据设计实现的Acro Media媒体服务器,不仅实现了媒体服务器的基本功能,并且提供大容量的语音会议资源和高清视频会议资源。Acro Media媒体服务器的插板扩充设计,满足NGN网络部署时对不同容量媒体资源的要求,具有高稳定性和可靠性。     

基于软交换的集群媒体服务器高可用性系统建模与分析

在研究基于软交换的集群媒体服务器的系统结构基础上,提出了一个高可用性系统方案。利用广义随机Petri网理论对高可用性方案进行了建模和仿真实验。实验结果表明,该方案可以显著提高系统的可用性,是一个有效方案。基于可用性模型分析了系统修复速率、请求处理机数量、以及请求处理机故障特征对集群可用性的影响,分析结论对基于软交换的集群媒体服务器的可用性设计具有重要参考价值。     

软交换中媒体服务器的研究与应用

软交换是下一代网络运营中的关键技术。主要论述了软交换中为各种业务提供媒体资源的基础组件——媒体服务器的功能和系统结构,并对其在软交换业务中的具体应用流程作了详尽叙述。     

IP媒体服务器控制协议的探讨

媒体服务器属于软交换网络中一个网元,在软交换或应用服务器的控制下,提供各种业务所需的媒体资源。是业务实现过程中不可或缺的组成部分,广泛应用于包括基本语音提示、IP Centrex、IP会议、预付费业务、统一通信、视频、IPFAX等各种业务。     

Resource management in software-programmable router operatingsystems

Future routers will not only forward data packets but also provide value-added services, such as security, accounting, caching, and resource management. These services ran be implemented as general programs, to be invoked by traversing packets embedding router program calls. Software-programmable routers pose new challenges in the design of router operating systems (OS). First, router programs will require access to diverse system resources. The resource demands of a large community of heterogeneous resource consumers must either be coordinated to enable cooperation or arbitrated to resolve competition. Second, it is beneficial to concurrently support multiple virtual machines, each with a guaranteed share of physical resources. This allows services to be customized and to seamlessly evolve. We present the design and implementation of a next generation router OS that can meet the above challenges. We define an orthogonal kernel abstraction of resource allocation, which can schedule various time-shared and space-shared resources with quality of service (QoS) differentiation and guarantees. A scalable and flexible packet classifier enables dynamic resource binding and per-flow processing of received packets. We have prototyped our system on a network of UltraSPARC and Pentium II computers. Currently, QoS-aware schedulers for CPU time, forwarding bandwidth, memory-store capacity, and capacity for secondary data stores have been integrated. We present experimental results on various aspects of resource management in our system     

如何引导大缓存

缓存是指在计算机存储系统的层次结构中,介于中央处理器和主存储器之间的高速小容量存储器。缓存和主存储器一起构成一级存储器,高速缓冲存储器和主存储器之间信息的调度和传送是由硬件自动进行。在计算机的发展历程中,依据摩尔定律,计算系统中的中央处理器性能发展迅速。而磁盘作为计算系统中的主存储器,由于机械机理的限制,其发展速度远远不及中央处理器的发展速度,形成了中央处理器数据处理快而磁盘读写数据缓慢的状况,从而降低整个计算机系统工作效率。因此,通过在两者之间增加一个缓冲层来协调两者之间数据调动效率问题,缓存由此应运而生。缓存的处理速率接近于中央处理器,可以通过扩大缓存容量,缓解两者之间处理效率差距,能够快速响应中央处理器和磁盘之间的读写请求,作为两者之间的缓冲池,缓存在一个适当范围内越大越好。由于缓存资源的珍贵,因此,缓存成为一个计算系统性能高低的重要标志。     

A Distributed and Elastic Application Processing Model for Mobile Cloud Computing

The latest developments in mobile computing technology have increased the computing capabilities of smart mobile devices (SMDs). However, SMDs are still constrained by low bandwidth, processing potential, storage capacity, and battery lifetime. To overcome these problems, the rich resources and powerful computational cloud is tapped for enabling intensive applications on SMDs. In Mobile Cloud Computing (MCC), application processing services of computational clouds are leveraged for alleviating resource limitations in SMDs. The particular deficiency of distributed architecture and runtime partitioning of the elastic mobile application are the challenging aspects of current offloading models. To address these issues of traditional models for computational offloading in MCC, this paper proposes a novel distributed and elastic applications processing (DEAP) model for intensive applications in MCC. We present an analytical model to evaluate the proposed DEAP model, and test a prototype application in the real MCC environment to demonstrate the usefulness of DEAP model. Computational offloading using the DEAP model minimizes resources utilization on SMD in the distributed processing of intensive mobile applications. Evaluation indicates a reduction of 74.6% in the overhead of runtime application partitioning and a 66.6% reduction in the CPU utilization for the execution of the application on SMD.

     

Architecture and design of NX-2700: a programmable single-chip HDTVall-format-decode-and-display processor

This paper describes the architecture, functionality, and design of NX-2700, a digital television and media processor chip from Philips Semiconductors. The NX-2700 is the second generation of an architectural family of programmable multimedia processors targeted at the digital television (DTV) markets, including the United States Advanced Television Systems Committee (ATSC) DTV-standard-based applications. The chip not only supports all of the 18 ATSC formats from standard-definition to wide-angle, high-definition video, but has also the power to handle high-definition television (HDTV) video and audio source decoding (high-level MPEG-5 AC-3 and ProLogic audio, closed captioning, etc.) as well as the flexibility to process advanced interactive services. NX-2700 is a programmable processor with a very powerful, general-purpose very long instruction word (VLIW) central processing unit (CPU) core that implements many nontrivial multimedia algorithms, coordinates all on-chip activities, and runs a small real-time operating system. The CPU core, aided by an array of peripheral devices (multimedia coprocessors and input-output units) and high-performance buses, facilitates concurrent processing of audio, video, graphics, and communication-data     

Microcosm TNG: a framework for distributed open hypermedia

Increased amounts of information and users' rising demands for advanced link services pose strong challenges to distributed hypermedia systems. Novel presentation facilities, nontextual digital media, personalization of content and structures, and many more techniques often require complex processing and thus imply new interaction styles. Microcosm TNG (The Next Generation), which is a framework for distributed open hypermedia, offers new mechanisms for desktop and Internet integration. It also promotes a modular approach to authoring hypermedia resources     

嵌入式WinCE多线程编程研究与实现

Windowsce5．0作为一个多任务操作系统,采用了一种新的任务调度机制。原理上它将一个进程划分为多个线程,每个线程按照一定的调度策略占用CPU的运行时间及其资源,这样使得CPU的调度单元很小,从而提高了CPU并发处理能力。本文以线程的创建,线程功能的实现,最后将2个线程在wince5．0上并发运行,达到了理论和实践的结合,更进一步了解了线程在嵌入式Windowsce中的调度机理     

一种支持硬件加速的虚拟网络功能部署模型

为解决以软件实现的虚拟网络功能(VNF)性能受限问题,软件定义网络和网络功能虚拟化(SDN/NFV)等新型网络架构引入了硬件加速资源。硬件加速资源的部署,使得VNF能够为日益增长的数据流量提供服务保障。该文针对已有研究未考虑具有高性能数据处理需求的服务链VNF部署问题,提出一种支持硬件加速的VNF部署模型。该模型基于硬件加速资源的承载特性,在保证未加速VNF到商用服务器的优化部署下,优先实现交换机中加速资源的复用,并根据网络业务的性能需求,灵活调整加速资源与VNF的映射约束。仿真实验表明,与其他典型部署方法相比,在引入相同硬件加速资源的情况下,该模型可以承载更多的业务流量,满足服务链高性能数据处理需求,有效提高了部署在网络中加速硬件的资源利用率。     

DSP芯片在实时图像处理系统中的应用

在通用计算机上用软件实现图像处理，要占用CPU几乎全部的处理能力，且速度相对较慢。数字信号处理器(DSP)的可编程性和强大的处理能力使其可用于快速实现各种数字信号处理算法，在图像处理领域，尤其在实时图像处理系统中得到了广泛应用和发展。本文介绍了DSP芯片及其在实时图像处理系统中的开发、应用及发展方向。     

基于NVIDIA GPU 的机载SAR 实时成像处理算法CUDA 设计与实现

合成孔径雷达(SAR)成像处理的运算量较大,在基于中央处理器(Central Processing Unit, CPU)的工作站或服务器上一般需要耗费较长的时间,无法满足实时性要求。借助于通用并行计算架构(CUDA)编程架构,该文提出一种基于图形处理器(GPU)的SAR 成像处理算法实现方案。该方案解决了GPU 显存不足以容纳一景SAR 数据时数据处理环节与内存/显存间数据传输环节的并行化问题,并能够支持多GPU 设备的并行处理,充分利用了GPU设备的计算资源。在NVIDIA K20C 和INTEL E5645 上的测试表明,与传统基于GPU 的SAR 成像处理算法相比,该方案能够达到数十倍的速度提升,显著降低了处理设备的功耗,提高了处理设备的便携性,能够达到每秒约36兆采样点的实时处理速度。      

Comparative study of two CPU router time management algorithms in cellular IP networks

Providing good quality of service (QoS) in cellular IP networks is an important requirement for performance improvement of the cellular IP network. Resource reservation is one of the methods used in achieving this goal and is proven to be effective. The main resources to be reserved in a cellular IP network are bandwidth, buffer and central processing unit (CPU) cycles. Router CPU cycle is the time taken by the router to process the packet of the flow before forwarding it to the next router (hop). This paper proposes a model for CPU cycle optimization of routers for real‐time flows in a cellular IP network. The model applies both genetic algorithm (GA) and particle swarm optimization (PSO) as soft computing tools to optimize the CPU cycles and reduces the flow processing time at each router in the route taken by a flow. Simulation experiments illustrate a comparative study of the model.