《LSF参数论文：基于条件矢量及内插量化的抗分组丢失方法》由会员分享，可在线阅读，更多相关《LSF参数论文：基于条件矢量及内插量化的抗分组丢失方法（3页珍藏版）》请在金锄头文库上搜索。
1、LSF参数论文：基于条件矢量及内插量化的抗分组丢失方法【中文摘要】在IP网络尤其是无线方式下因信道资源受限造成 的实时语音的低速率传输和数据丢包现象难以避免。边信息(CSI)编码是节省数据量和接收端丢包信号恢复的有效途径。如何有效利用边 信息,用尽可能少的比特承载和恢复高质量语音是 CSI的关键技术。 本文基于相关信源数据关联特点，为实现质量可控的量化编码，利用 线谱频率(LSF)参数的帧间帧内边信息设计了不同精度的条件矢量量 化(CVQ)码书，提出了 CVC与相邻帧间参数内插相结合的 CVQ-1算法， 应用于G.729编码;结合交织技术的误差分散优点，形成交织的 CVQ-II-G.729方案
2、。本文主要工作如下：1、提出LSF系数的CVQ-I 算法。首先对帧内LSF系数低、高频分量分别用LBG算法构建VQ码 书。由统计条件矢量分布生成两者的关联码书，其关联码字的生成准 则是对应LSF估计值均方误差。当LSF参数丢失时，利用帧内量化的 高频LSF码字和相邻帧间低频LSF码字作为边信息，估计重要性较高 的低频LSF参数。实验证明，考虑到码书存储空间、带宽和恢复质量 的折中，冗余比特选取为3bits时恢复的LSF较好。LSF参数丢失10% 的情况下,CVQ-I方法恢复的LSF参数失真较内插方法失真降低 0.0076dB。2、应用CVQ-I原理设计不同精度的高、低频 LSF参数VQ 码书，
3、并对其构造相应的关联CVQ码书。设计了四种可变速率的联合 码书来估计低频LSF参数,满足不同LSF估计精度和带宽的要求。3、 应用CVQ-I改进了 G.729丢失帧隐藏方法。将一帧内LSF参数低频保留，高频交织到相邻帧，实现了 CVQ-II丢失帧容错方法。实验证明，相比G.729的帧隐藏方法，随机信道丢包下的PES兪内插隐藏方法平 均提高0.03。【英文摘要】 Low rate compressi on and packet loss inreal-time VOIP(Voice over IP) tran simissi onis in evitabledueto the limited c
4、ha nnel resource especially un der wireless channels. Coding with side information (CSI) is an effective way to reduce data and recover the lost signal. At the receiving end, it is the key point to describe the high quality speech by utiliz ing CSI efficie ntly with fewer bits .In this thesis, diffe
5、rentconditionalvector quantization(CVQ) codebooks aredesig ned to realize the accuracy con siste ntLSF qua ntizati on,the quality con trol is also fulfilled by the related codebook associated with the CSI of i ntra-frame and in ter-frame Lin ear Spectrum Freque ncy (LSF). A new CVQ scheme with lin e
6、ar in terpolati on (CVQ-I) is suggested in G.729. Furthermore, an in terleaved CVQ-II for packet lost con cealme nt is also give n.Our main work is focus on:1. Estimation of the lost LSFsscheme-CVQ-I method. The in tra-frameLSFs are first splittedinto two parts, the low frequency component (LFC) and
7、 the high one (HFC), each are quantized by its VQcodebook trained by LBGclustering.A related codebook is created based on conditionaldistributions of concerning the LFC and HFC by minimizing the errors of LSF estimation.A CSI prediction with intra-frameHFCand inter-frame LFC are providing as side in
8、formation to estimate the lost LFC. Experime nts prove that CVQ-I with 3 redundancy bits (CVQ-I3) is better for the compromise of the storage space of codebook, ban dwidth and recovery. At ran dom loss rate 10%, the SDof our CVQ-I3 is 0.0076dB below than that of the linear interpolation (Ll).2. Accu
9、racy consistent codebook design. Different accuracy LFC, HFCand the correlated CVQ codebooks are desig ned by CVQ-I method. Four rates of the three codebooks are desig ned to meet differe nt estimated accuracy and ban dwidth.3. A lost con cealme nt forG.729-CVQ-II. The in tra-frame HFC are in terlea
10、ved to the consecutive frame, thus the lost LFCcan be estimated by CVQ-I. The Experiment proves that at random loss rate 10% CVQ-II is 0.03 higher in PESQ than interpolation concealment (CI).【关键词】LSF参数 条件矢量量化(CVQ)交织G.729Iff【备注】索购全文在线加好友 区：139.938848同时提供论文写作一对一指导和论文发表委托服务【英文关键词】Lin ear spectrum Freque ncy(LSF)Con diti onal vector qua ntizati on (CVQ)In terleave
