适于ATM技术的可变速率语音编码

文章介绍一种适于结合ATM技术的高级专用网络的可变速率嵌入式ADPCM语音编码方案，目的是即使在这些网络内发生高比例信元丢失的情况下也能保特高质量的语音。     

ATM网络中恒定比特率和可变比特率语音通信

分析了ＡＴＭ（异步转移模式）网络中恒定比特率（ＣＢＲ）和可变比特率（ＶＢＲ）语音通信方式，讨论了ＩＴＵ－Ｔ１９９７年新颁布的ＡＴＭ可变比特率业务适配标准ＡＡＬ２，给出了信元装配实例。针对ＶＢＲ业务相对ＣＢＲ业务有较大不同的网络传输时延特性，着重分析了语音信号作为ＶＢＲ业务采用ＡＡＬ２适配标准时，通过ＡＴＭ网络传输时的装配时延问题。     

混合型中速率语音编码系统

设计了一种新型的中速率混合语压编码系统，该系统把语音分割成基带（０．３～１ｋＨｚ）和高频部分（１～３．４ｋＨｚ）对于重要的基带信号，采用高质量的４ｂｉｔ／样点的ＡＤＰＣＭ技术；对于相对次要的高频信号，采用高效的ＶＱ（矢量量化）技术，以压缩码率，对于矢量量化，还提出了一种新的快速算法，通过某种预处理使得搜索码本的速度提高１０倍以上，且质量等效于全搜索方法，本系统具有实现简单，时延短的特点，且主观质量     

用AAL5实现基于ATM的分组语音通信

本分析了现有应用于ATM分组语音通信的AAL1、AAL2适配方案优缺点后，提出了用AAL5实现基于ATM的分组语音通信，并就语音编码方式和时延这两个关键问题进行了探讨。     

ATM网络的视频编码技术

异步转换模式（ＡＴＭ）的提出对编码技术产生了重大的影响。本文概述了ＡＴＭ下可变比特率（ＶＢＲ）视频编码的优点及存在的问题，扼要介绍了各种信元丢失的预防及补偿方法，着重分析了分层编码技术，并对目前提出的各种可变比特率分层编码算法进行了概括性的评述，指出了有待研究的一些问题。     

ATM网络的视频编码

异步转移模式概念的提出对编码技术产生了重大的影响。本文概述了ＡＴＭ不可变比特率视频编码的优点及问题，扼要介绍了各种信元丢失的保护方法，着重分析了分层编码技术，并对目前提出的各种可变比特率分层编码算法进行了概括性的评述，指出有待研究的一些问题。     

An integrated FEC coding scheme for ATM transmission over regenerative satellite networks

The application of asynchronous transfer mode (ATM) on both wireless and satellite networks requires system adaptation. This adaptation has to improve the overall system's performance, and achieve high quality‐of‐service classes approaching that for fibre‐optic communications. In this paper, a new integrated forward‐error‐correction (FEC) coding scheme is introduced for ATM transmission over regenerative satellite networks. The proposed FEC scheme is a concatenation of two Reed–Solomon codes tailored for the header and payload parts of the ATM cell. This integrated coding scheme is shown to significantly improve the cell loss ratio as compared to the standard CRC code used in the ATM cell header. We obtain both upper and lower performance bounds for the concatenated code and check their accuracy when compared to exact system's performance. Both analytical and simulation results show that a cell loss ratio and bit‐error rate (BER) of 10^?25 and 10^?7 can be, respectively, achieved with minimum delay requirements on the SATCOM link. Finally, an approximation for the system's throughout is obtained. It is shown that using a hybrid selective‐repeat automatic‐repeat‐request (SR‐ARQ) with the RS code, a large throughput of approximately 0.843 can be achieved at BERs lower than 10^?7 for data services. Copyright © 2005 John Wiley & Sons, Ltd.     

IP和ATM网络中语音分组丢弃对解码语音质量的影响及对应的措施

基于分组传输的网络(如IP和ATM)，进行实时语音通信时，存在分组可能丢失，因而影响通信质量的问题。本文提出两种新的方法，以减少或降低语音分组丢失带来的影响。实验结果表明，它们明显改善了语音通信质量。     

VIDEO CODING SCHEME FOR ATM NETWORKS

A new variable bit rate layered coding scheme based on subband coding is given in this paper. It can make full use of the merits of ATM transmission, as well as compensation for the quality degradation caused by cell loss, which is an inherent problem of ATM networks, and thus a stable, high quality video transmission is achieved. This paper presents the coding scheme, cell construction and simulation results.     

语音业务在ATM网络中降低装配时延提高系统质量的新方法

本文针对ＡＴＭ网络中的语音通信，提出了一种降低信元装配时延的新方法，以及相应的基于ＡＤＰＣＭ的新的信源编码系统，它能确保在信元丢失时，收发双方编译码系统不会失步，并且能产生基本无附加听觉噪声的丢失段恢复语音。文中还分析了各部分参数的求法，实验表明，该法产生的丢失段恢复语音，具有良好的听觉质量。     

基于网络编码的低信噪比多跳中继传输方案

针对传统的反馈机制在低信噪比多跳中继通信中时延较高的问题,提出了一种适用于低信噪比场景下多跳中继通信的编码传输方案.该方案在已构建的多跳中继模型上,通过马尔科夫链分析传输时间与编码参数的关系,并结合信道条件优化编码参数,以获取较短的传输时间.在编码参数的优化过程中,提出了"ε初始"搜索算法.该算法调整搜索的顺序,以提高...     

基于动态小波神经网络非线性预测的语音编码系统

现在,所有的语音编码系统都采用线性预测技术,但对于本质非线性的语音信号而言,线性预测是不够的.因此,本文提出一种带反馈单元的动态小波神经网络并将其应用于语音编码系统,并对其函数逼近能力和学习高维函数的优越性进行分析.由于反馈单元的内部记忆能力,动态神经网络具有对长时相关的预测能力并能在一定程度上克服小波神经网络的"维数灾难"问题;在对语音信号的预测中,动态小波神经网络预测器的预测性能很好,虽然其预测阶数很低(仅为2).由于预测器较好的预测性能,当将此预测器用于语音编码系统中的后向预测时,实验结果表明:新系统的恢复语音平均分段信噪比比ITU的G.721标准提高3～4dB但二者码率相同.另外非线性预测语音编码系统的计算量是可以接受的.     

无线ATM通信网中一种新的快速越区切换方案

本文根据地无线ATM通信网中快速越区切换问题的研究,提出了一种新的基于虚道路（VP）的快速越区切换方案,并通过建立简化的呼叫模型对该方案的切换性能进行了理论分析和比较,其结果表明,该方案的各项性能均比虚拟连接树（VCT）方案和永久虚连接（PVC）方案优越。     

VBR wavelet video coding and its rate control in ATM networks

This paper presents a new video coding system based on wavelet transform and its rate control scheme over ATM networks. First, three dimensional wavelet transform is performed for the original image sequence, and an extension of set partitioning in hierarchical trees algorithm is employed to quantize the wavelet coefficients. Then, the output rate of the coder is controlled at group of frame scale, ensuring that it conforms to the parameters of a leaky bucket controller. Several leaky buckets with different sizes are discussed too. Simulation shows the efficiency of this codec and the effectiveness of the proposed rate control scheme.     

A dynamic transmission capacity allocation scheme for wired/wireless ATM networks

The previous research on administration of the transmission capacity in the wired/ wireless ATM networks only focuses on wired part or wireless part. There are very few people do the work extending to the links associating with handoff in the whole network. This paper develops the algorithms of transmission capacity administration on the link connecting the base station and base station controller (including the air interface of the base station) and the VPs among the base station controllers in the wired/ wireless ATM networks, which adapt to the traffic state of each service in every cellular cell to allocate (provision) transmission capacity and to reserve handoff guard capacity on these links, respectively. By simulating and analyzing the performance of the algorithm, it is found that it does well for the multimedia communication in which the transmission capacity requirement of each service may be widely different, so that the network bandwidth resource can be used efficiently.     

一种基于回归神经网络的码本激励非线性预测话音编码算法

为改善预测类声码器中长时预测器特性，本文引入了一种全连接回归神经网络（ＦＲＮＮ）非线性预测器并将其应用于话音编码算法中。ＦＲＮＮ在隐层单元不仅有来自自身的反馈，也有来自输出单元的反馈，因此其预测性能好于常规预测器。将其应用于码本激励话音编码系统（ＣＥＬＰ）中，可得到较低的传输码率，同时亦可改善编码质量。