一种基于聚合流的端到端统计时延界的新算法

为分析融合网络中聚合业务的端到端时延性能,提出一种基于聚合流的融合网络端到端统计时延界的新算法.该算法利用MGF(矩母函数)重新表征了网络端到端时延界的MGF形式的概率模型.数值分析结果表明了该算法的有效性和优越性,该算法很大程度上提高了独立统计复用,对融合网络性能评价具有参考意义.     

IMS下的固定和移动网络融合架构探讨

IMS下的网络融合作为当前通讯业界所关注的热点议题,能有效实现其固定和移动网络融合,对于推动通讯业界发展具有关键性作用,本文通过对IMS作为固定和移动业务网络融合平台的优势环节进行分析,对IMS的业务生成环境进行介绍,并从中提出相应IMS业务网络融合架构的创想.     

多业务融合承载网与传输网的关键技术分析

IP化是未来电信网络发展的必然趋势.业内越来越关注IP承载技术和传输技术的分工和融合,作为网络层面融合的主旋律,IP网络必将成为未来语音、视频、数据以及各种综合多媒体业务统一承载的网络平台,这在电信界已经达成共识.但是,作为多业务融合承载网络,传统的IP网络很难满足电信级业务对QoS、安全、可靠性和管控的要求.因此,需要对传统IP网络进行改进和优化,即保留IP灵活方便的特性,又满足电信业务承载和运营商对网络的新要求.     

基于IMS的固定和移动网络融合

3GPP提出的IP多媒体子系统（IMS）技术体现了通信网和因特网技术的结合、下一代网络和3G网络技术的结合,已受到通信界和信息技术界的广泛认同,成为固定和移动网络触合的核心技术。其网络融合综合了互联网IP技术、软交换技术和蜂窝核心网技术,业务融合综合了开放式Parlay／OSA技术和Web Services技术,管理触合综合了电信网管理、IP网络管理和策略控制技术。面对互联网的严峻挑战,各个标准化组织正通力合作推进IMS的发展,统一IMS、业务代理和有管理的P2P业务是进一步研究的重要内容。     

多业务融合承载网与传输网的关键技术分析

IP化是未来电信网络发展的必然趋势。业内越来越关注IP承载技术和传输技术的分工和融合，作为网络层面融合的主旋律，IP网络必将成为未来语音、视频、数据以及各种综合多媒体业务统一承载的网络平台，这在电信界已经达成共识。但是，作为多业务融合承载网络，传统的IP网络很难满足电信级业务对QoS、安全、可靠性和管控的要求。因此，需要对传统IP网络进行改进和优化，即保留IP灵活方便的特性，又满足电信业务承载和运营商对网络的新要求。[编者按]     

固定移动网络融合策略探讨

固定移动融合成为当今电信界的热点问题,结合现有的网络情况,探索未来的融合之路,也是当前运营商关注的问题,但目前对固定移动融合具体的实施策略还没有完整统一的规划。文章从不同侧面分析了固定移动网络融合的相关实施策略,包括从不同类型运营商角度以及网络不同层面角度进行分析,并提出了相关建议,同时分析了实施固定移动融合对网络运营模式的影响。电信运营商需要结合自身网络结构、运维体制、业务运营模式等作出规划分析,找出适合自己的网络融合之道。     

IMS的发展现状和应用综述

随着通信技术和信息应用的发展,电信业务需求正在向固定移动融合、语音与互联网应用相结合的综合信息服务转变,传统运营商面临移动运营商与互联网应用提供商的激烈竞争。IMS（IP多媒体子系统）作为下一代固定和移动网络融合核心技术正引起通信界和信息技术界广泛的认同。本文分析了IMS的技术现状及进展情况、IMS融合技术以及IMS发展所面临的问题等。     

下一代互联网研究方向与趋势

"业务综合"和"网络融合"是通信界多年来努力的目标。IP网络具有承载包括语音、数据和视频等多种业务应用能力,这使业界将全IP网络看作是未来融合网络的雏形。但目前的IP网络无法保证实时业务的服务质量以及网络和业务的安全性。为了在下一代互联网发展方面处于领先地位,全球各国都在不断投入力量进行下一代互联网的研究工作。本文主要介绍IP网络和业务的发展情况、目前IP网络存在的问题以及全球在发展下一代互联网方面所做的努力。     

支持网络融合的策略管理系统研究

网络融合为网络控制管理引入了较多的复杂性,策略管理控制是解决这一问题的关键技术。已有的策略管理系统架构具有扩展性不好、自动化程度不高、动态适应性较弱等缺点。在分析网络融合对于策略管理系统需求的基础上,采用面向服务架构,设计了策略管理系统框架,使其能方便的适应异构网络技术、多样化业务与复杂多变的应用环境。针对NGN等主流网络融合技术的特点,提出了网络管理控制的策略体系。策略管理系统框架能够较好的支持网络融合的环境。     

军民融合的通信网络构建

军民融合是我国发展社会经济和国防建设的重要战略,其目的是打破原有的资源利用界限,实现"一份投入,两份产出",而军民通信融合是军民融合中的重要组成.本文以军民融合通信网络的构建作为研究核心,提出了两种军民融合通信网络的模式,并给出了这两种军民融合通信网络模式所对应的构建方案,希望能为我国军民融合通信网络的构建和完善提供一定的参考意义.     

Broadband network restoration

In broadband networks, it is expected that at a given time there will be more data in the network than in narrowband networks, and therefore, to maintain the same quality of service, failures will need to be restored much faster. The purpose of this article is to demonstrate that broadband network technology, and in particular the asynchronous transfer mode (ATM), has factors that enable much faster restoration in broadband networks. These can be summarized as: (i) the higher-speed processing and larger-volume storage possible with today's technology as compared to what was available for narrowband networks, (ii) faster failure detection by making use of ATM cell header error checks, (iii) easier and faster rerouting simply by changing cell headers, (iv) easier and faster splitting of virtual paths by changing cell headers, (v) higher fill factors in the new digital hierarchy due to full termination at cross-connects and switches, and (vi) the statistical multiplexing advantage that can be exploited in a full ATM network configuration. The authors describe four basic characteristics of broadband network restoration, made faster and more efficient because of the factors listed above. These are (i) hitless protection switching, (ii) robust and fast failure detection, (iii) restoration alternatives in broadband networks, and (iv) fast topology update for multiple failures. By employing a combination of these methods, fast restoration of broadband networks can be achieved. The techniques described enable a B-ISDN/ATM network to be constructed that is highly self-healing (i.e., can recover from failures without human intervention) and very fast     

Discrete-time dynamic image segmentation system

A dynamic image segmentation system that consists of a chaotic neural network and a global inhibitor has been developed. this is a discrete-time system and is faster than a locally excitatory globally inhibitory network (LEGION) that is a similar system with continuous-time dynamics. it is demonstrated that the system works well and is faster than the LEGION.     

Approximation multidimension fuction with functional network

The functional network was introduced by E.Catillo, which extended the neural network. Not only can it solve the problems solved, but also it can formulate the ones that cannot be solved by traditional network. This paper applies functional network to approximate the multidimension function under the ridgelet theory. The method performs more stable and faster than the traditional neural network. The numerical examples demonstrate the performance.     

一种最小化编码节点的网络编码优化算法

网络编码能有效地提升多播网络的传输性能,但编码的引入增加了节点的计算开销。为了克服网络编码带来的额外开销,该文提出了在代数网络编码框架下的网络编码优化模型,并在此模型基础上给出了基于改进遗传算法的最小化编码节点算法-(MCN,Minimizing Coding Nodes)。MCN在简单遗传算法的基础上增加了一些新的策略,避免了局部性问题和降低了算法寻优时间。模拟实验结果表明,MCN是有效的而且运行的更快,输出的网络编码方案所需要的编码节点也更少。同时将MCN应用到具有实际意义的网络中,同传统的网络编码相比,吞吐率仍可达到25%以上,而网络的平均延迟和网络开销却大大减少。     

Flexible and Expandable Speech Recognition Hardware with Weighted Finite State Transducers

Hardware implementation of speech recognition can not only accelerate decoding speed for real-time processing but also reduce the power consumption. Recently the weighted finite state transducer (WFST) has emerged as a major recognition network representation because it reduces the algorithmic complexity of decoding procedures by applying many optimizations on the network in offline. However, hardware implementation of continuous speech recognition (CSR) with the WFST network is challenging, mainly because Viterbi search should traverse a large sized network with limited hardware resources. This paper presents two hardware speech recognition systems with the WFST network. The first one, which is called the SRAM-oriented system, utilizes the internal SRAM as a hash table to efficiently manage active working set. This system is flexible because it can easily accommodate different speech recognition tasks as long as the SRAM space is allowed. For easy expansion, we also propose the DRAM-oriented system where the active working set is stored in the external DRAM. To hide long latency of DRAM access, a split DRAM hash table is employed, which stores active working set in the opened rows of DRAM to reduce the number of row misses. Experimental results show that the SRAM-oriented system decodes the 5k-word CSR task 4.93 times faster than real-time, while the DRAM-oriented system runs 4.48 times faster than real-time with only about a half SRAM capacity.     

用Hopfield网络实现盲波束形成

本文提出了一种基于改进Ｈｏｐｆｉｅｌｄ 网络的盲波束形成算法。利用神经网络的计算结构,避免了矩阵求逆运算,同时利用对角加载技术加快了收敛速度、保证了算法的鲁棒性     

LFM-M码信号旁瓣抑制技术

为了进一步提高雷达的探测性能,设计了线性调频–二相码(LFM-M)混合调制脉冲压缩信号。采用分类比较的方法,研究了反向传播网络、Elman网络和径向基函数(RBF)网络等3种典型神经网络在其脉冲压缩中的应用,设计了网络的结构,分析了网络的算法。通过仿真和对脉冲压缩输出性能的研究得出,采用RBF神经网络对LFM-M码信号进行脉冲压缩,网络具有较快的收敛速度和较好的数值稳定性,可获得60 d B左右的输出主旁瓣比。     

基于UDP的无线自相似环境下MPEG-4码率自适应控制

对在无线局域网环境下，基于UDP的多媒体流码率自适应控制进行了研究，同时探讨了网络自相似等相关问题，并借鉴通信原理中的自适应滤波方法，解决网络较平稳时码率收敛问题。通过在实验环境下对理论结果加以测试验证，实现了在自相似、多级分形的无线网络状况下，RTP／UDP／IP协议的多媒体码率自适应控制。该方案能更有效地跟踪实时的网络有效带宽，比传统模型的方案更快速、准确，保证服务质量。特别地，采用带收敛特性的自适应算法后，当网络有效带宽趋于稳定时，媒体流的输出速率也收敛并趋于稳定，有利于server和网络效率的改善。     

改进BP神经网络在物体识别中的应用

针对传统BP神经网络容易陷入局部极小、收敛速度慢和确定隐含层的神经元个数比较困难等缺点,从结构和算法两方面对BP神经网络进行改进。改进后的网络具有较快的收敛速度和较短的运行时间,加强了BP神经网络的学习能力和自适应能力,并将其应用于物体的分类识别,取得了良好的效果。仿真结果表明了此改进方法的可行性和有效性。     

Energy-efficient utility maximization for wireless networks with/without multipath routing

In wireless networks, end-to-end communication depends on link capacities which, in turn, are determined by transmit powers of interfering links. Optimal network performance and energy efficiency can be achieved by jointly optimizing congestion control and power control. In this paper, we study this joint optimization problem by formulating it into convex programming, i.e., we maximize a compound function which is a network utility function minus a factor, named tradeoff factor, of the associated power cost. We prove that this tradeoff factor is essential for good energy efficiency while maintaining the network throughput at a satisfactory level. The problem is solved by a distributed dual-decomposition based algorithm energy efficient jointly optimal congestion and power control (EJOC). EJOC tackles the power control problem in a recursive manner, operating as easily as the steepest descent method but converging much faster. This optimization framework is further extended to networks where each data source may have multiple alternative paths to its destination. Simulation results show that the proposed algorithm converges faster than other algorithm and is capable of significantly improving the energy efficiency of the network.