电子信息与通信专业外语(5)

(4) Coding

Coding, that is the expression of a quantized sample magnitude in terms of a binary number for transmission over the channel, may be executed in many ways. The main categories may be described as:

Parallel coding

Sequential coding

Counting.

The basis of counting methods is to count the successive levels on a simple binary counter at an accurately controlled rate of count and simultaneously to develop a steadily rising potential

部分章节译文

( sawtooth wave form). If, when this potential matches that of the sample, the counting is arrested the desired coding has been achieved.

An interesting version of this method comprises the replacement of the linear (sawtooth) wave form by an exponential decay of the sample, the count being arrested when a predetermined limit is reached . This gives effectively logarithmic compression.

Counting methods such as these have the merit of being monotonic and although there may be inaccuracies due to deviations in the slope of the sawtooth, these will be consistent from sample to sample. The sequential methods to be described next introduce greater risks of irregularities and discontinuities in the shape of the companding curve.

The main objection to counting methods is that unless very high speed counters are used the time involved in step-by-step counting of samples involving a large number of levels is excessive.

The procedure which is most convenient for the handling of PCM voice is the sequential method or the development of the code digit by digit. Fig. 3.6 indicates the essential nature of this method. The sample is compared digit by digit with a potential derived from a weighting network representing the cumulative total as the number is built up. Each decision is final and requires no subsequent modification, so that digits may be transmitted to line as the coding proceeds. Further, it is relatively simple to make the weight assigned to say the last three digits a function of the preceding interim total, which leads to easy implementation of the segmented approximation to the companding law. This is particularly simple if the ratio of the slopes of consecutive segments

is an integral number.

Fig.3.6 Principle of sequential encoding ( linear )

This method of coding and the use of a segmented curve leads to the adoption of symmetrical binary as the code structure. In thus , one digit defines the sign of the sample in regard to a selected centre point and the remaining digits define in a straight binary manner the magnitude of the sample +ve or - ve regardless of sign.

In current specifications the aim has been to achieve performance standards which are satisfactory not only for one link but for a number of consecutive links with interconnection at audio level . This process must share with f.d.m. the inevitable degradation of the frequency/loss curve due to consecutive filters but it also means that the quantising and idle channel noise will be additive. The effect in regard to quantising noise is substantially that of r.m.s. addition. Idle channel noise is rather more complicated to deal with. It must inescapably be responsive to the presence or absence of loss in the audio interconnection and to the pedestal relationship prevailing at each stage. In general the build-up is rather less than straight r.m.s. addition and low frequency modulation by hum tends to become less prominent.

部分章节译文

脉冲编码调制与数字传输

（一）简介

一个模拟波形的转换过程，如从电话语音到数字形式不可避免的包含一个带宽，或在使用介质所占用的频谱中最先出现一个相对没有意义的复杂的传送语音信号的过程。在技术和经济上的优势将会在下面的章节解释，一个简要的轮廓可以帮助解释这个观点。

多年来人们尝试实现时分多址的经济潜力。这就意味着有一个新概念，电话的实现在电报使用之前的。所应用的设备只是同步交换机，在发送端顺序地对每一个通道采样，并顺序地将这些采样值发送出去。另一些接收端的设备接收采样值，并对每个通信信道的采样值按照正确的顺序分配。显然在同步交换机的维护方面有问题。这个问题是现代网络已经提供廉价的、可靠的方式来获得需要的结果。

正如在电报中，信息是数字的，包含一系列的信息单元限制一些不连续值，例如，二进制符号与电报空间，时分多址是极其简单的并且廉价的。

长期以来人们一直都了解，其中有限带宽模拟波形的最高频率为f1，这个频率可准确地传送，必须要用瞬时采样值略微超过2f1的采样速率。用这些采样值作为窄脉冲传输，在对应几个不同信道的脉冲可以用一个整周期内连续地送出去，使时分多路复用技术应用于模拟信息实现成为可能。

时分多路复用以传输为目的，也用于转换，各种调制技术已经被采用：例如，脉冲幅度，脉冲宽度和脉冲位移。在切换过程中，采样值可以在宽带上发送，可以达到合理的控制信号损伤，但在连接点和干线中的幅度和相位脉冲的必然失真，使其在没有均衡时，很难控制信道间干扰或串扰。

时分多路复用过程的中产生问题，与对早期无线链路高噪音水平相反，已故的亚历克里夫斯在30年前就构思了多路复用，对其中一个模拟波形进行采样，定期由窄脉冲和每个抽样的振幅数值来“说明”替代模拟波形的传输。

虽然最初的构想是，大规模商用的脉冲编码调制用在无线电噪声中，而在双电缆中使用脉冲编码调制。随着经济的不断增长，大量的安装在短的、中间的双铜线通道已经被识别。这些具有相对较差串扰和噪声特性的双绞线，已经证明了通过传统的频分多路复用引入的复用这个阻碍。脉冲编码调制已经证明可行，这将在以下章节中来解释，用两个规则的双绞线处理从24路到32路的会话。

百度搜索“77cn”或“免费范文网”即可找到本站免费阅读全部范文。收藏本站方便下次阅读，免费范文网，提供经典小说公务员考试电子信息与通信专业外语(5)在线全文阅读。