汉英对照:
Chinese-English Translation:
群同步的实现方法通常有两类：一是在数字信息流中插入一些特殊码组作为每群的头尾标记，接收端根据这些特殊码组的位置实现群同步；另一类是利用数据码组本身之间彼此不同的特性来实现自同步。
There are two kinds of group synchronization methods: one is to insert some special code groups into the digital information stream as the head and tail marks of each group, and the receiver realizes group synchronization according to the position of these special code groups; the other is to realize self synchronization by using the different characteristics of data code groups.
在群同步的通信系统中，传输的信息被分成若干”群”。所谓的”群”,一般是以字符为单位，在每个字符的前面冠以起始位、结束处加上终止位，从而组成一个字符序列。数据传输过程中，字符可顺序出现在比特流中，字符与字符间的间隔时间是任意的，即字符间采用异步定时，但字符中的各个比特用固定的时钟频率传输。群同步的目的是将接收码元正确分组。
In group synchronous communication system, the transmitted information is divided into several “groups”. The so-called “group” usually takes characters as a unit, and each character is preceded by a start bit and a stop bit is added at the end to form a character sequence. In the process of data transmission, characters can appear in the bitstream in sequence, and the interval between characters is arbitrary, that is, asynchronous timing is used between characters, but each bit in the character is transmitted at a fixed clock frequency. The purpose of group synchronization is to correctly group received symbols.
群同步问题实质上是一个对群同步标志进行检测的问题。 对群同步系统的基本要求是：
Group synchronization is essentially a problem of detecting group synchronization flag. The basic requirements of group synchronization system are as follows:
1、正确建立同步的概率要大， 即漏同步概率要小， 错误同步或假同步的概率要小。
1. The probability of correctly establishing synchronization is large, that is, the probability of missing synchronization is small, and the probability of false synchronization or false synchronization is small.
2、捕获时间要短， 即同步建立的时间要短。
2. The capture time should be short, that is, the synchronization establishment time should be short.
3、稳定地保持同步。 采取保持措施， 使同步保持时间持久稳定。
3. Keep in sync steadily. Take measures to keep the synchronization time stable.
4、在满足群同步性能要求条件下，群同步码的长度应尽可能短些，这样可以提高信息传输效率。
4. In this way, the transmission efficiency of group synchronization code can be improved as much as possible.
群同步的方法
Methods of group synchronization
群同步的实现方法通常有两类：一是在数字信息流中插入一些特殊码组作为每群的头尾标记，接收端根据这些特殊码组的位置实现群同步；另一类是利用数据码组本身之间彼此不同的特性来实现自同步。
There are two kinds of group synchronization methods: one is to insert some special code groups into the digital information stream as the head and tail marks of each group, and the receiver realizes group synchronization according to the position of these special code groups; the other is to realize self synchronization by using the different characteristics of data code groups.
起止式同步法
Start stop synchronous method
这是电传机中广泛使用的方法。电传报的一个码字由 7.5 个码元组成，它以一个码元宽度的负脉冲作为起脉冲，以 1.5 码元宽度的正脉冲作为止脉冲，其余 5 个码元表示有用的信息。接收端就根据 1.5 个码元宽度的正电平第一次转换到负电平这一特殊规律，确定一个码字的起始位置，因而实现了群同步。
This is a method widely used in telex. A code word of telex message is composed of 7.5 symbols. It takes a negative pulse of one symbol width as the starting pulse, a positive pulse of 1.5 code element width as the stop pulse, and the other five symbols represent useful information. According to the special law that the positive level of 1.5 symbol width is converted to the negative level for the first time, the receiver determines the starting position of a code word, thus realizing group synchronization.
缺点：
Disadvantages:
1、由于止脉冲宽度与码元宽度不一致，会给同步数字传输带来不便；
1. Because the stop pulse width is different from the symbol width, it will bring inconvenience to the synchronous digital transmission;
2、7.5 个码元中只有 5 个码元用于传输信息，效率较低；
2. Only 5 of the 7.5 symbols are used to transmit information, so the efficiency is low;
连贯式插入法
Continuous insertion method
这种方法是用特殊字符建立群同步，而且是在每群的开始集中插入群同步码组，作为群同步用的特殊码组要求有尖锐单峰特性的局部自相关函数，且识别器尽量简单。最常见的群同步码组是巴克码（Barker），它是一种具有特殊规律的二进制码组，是有限长的非周期序列。
This method uses special characters to set up group synchronization, and inserts group synchronization code group at the beginning of each group. As a special code group for group synchronization, it requires a local autocorrelation function with sharp unimodal characteristics, and the recognizer is as simple as possible. The most common group synchronization code group is Barker code, which is a kind of binary code group with special rules and is a finite length aperiodic sequence.
巴克码的局部自相关特性有一个假设前提：相邻码元为 0！实际上，当相邻的随机序列等概率取±1 时，巴克码近似有如此的相关特性；但对于任意的二进制数据，巴克码太短，并不能在所有情况下做最佳相关码的近似。为此，Willard 利用计算机仿真，找到了适于随机相邻码元，具有与巴克码相同长度、使假同步概率最小的 Willard 序列。
The local autocorrelation property of Barker code has a hypothesis that the adjacent code is 0! In fact, when the equal probability of adjacent random sequence is ± 1, Barker code approximation has such correlation characteristics; but for arbitrary binary data, Barker code is too short to approximate the best correlation code in all cases. Therefore, Willard uses computer simulation to find a Willard sequence with the same length as Barker code and minimum false synchronization probability.
连贯插入同步码是一个码组，要使同步可靠，同步码组就要有一定的长度，从而降低了传输效率。而分散插入则是每帧只插入一位作为帧同步码。例如北美和日本采用的 24 路 PCM，每帧有 8*24=192 信息码元，每逢奇帧其后插入一位帧同步码，1010…交替插入。由于每帧只插入一位，它和信息码元混淆的概率为 1/2，这样似乎无法识别帧同步码。不过分散插入方式在捕获同步时，并不是只检测 1 帧 2 帧，而是要连续检测 10 帧以上，每帧都符合“1”、“0”交替的规律才确认同步。误同步概率是很小的。
Coherent insertion synchronization code is a code group. In order to make the synchronization reliable, the synchronization code group must have a certain length, thus reducing the transmission efficiency. In the decentralized insertion, only one bit is inserted into each frame as the frame synchronization code. For example, in North America and Japan, there are 8 * 24 = 192 information symbols in each frame, and a bit frame synchronization code is inserted after every odd frame Insert alternately. Since only one bit is inserted into each frame, the probability of confusion between it and information symbol is 1 / 2, so it seems that the frame synchronization code cannot be recognized. However, when the decentralized insertion method is used to capture synchronization, it is not only to detect 1 frame and 2 frames, but to detect more than 10 frames continuously, and each frame conforms to the rule of alternating “1” and “0”. The probability of false synchronization is very small.
分散插入法传输效率高，但同步捕获时间长。
The distributed insertion method has high transmission efficiency, but the synchronization acquisition time is long.
群同步与位同步的区别与联系
The difference and relation between group synchronization and bit synchronization
载波同步解决了同步解调问题，把频带信号解调为基带信号，而位同步确定各码元的抽样判决时刻，即区分每个码元，使接收端得到一连串的码元序列，这串码元序列代表一定的信息。通常由若干个码元代表一个字母（或符号、数字），而由若干个字母组成一个字，若干个字组成一个句子。在传输数据时，若干个码元会组成一个码组。群同步的任务是把字、句或码组区分出来。例如，一个字由 20 个码元组成，则将码元位同步脉冲频率 fs 进行 20 次分频，即可得到字的定时脉冲频率。但每个群的“开头”和“结尾”的时刻，即群同步脉冲的相位是不能直接从同步脉冲中得到的，需要群同步解决。
Carrier synchronization solves the problem of synchronous demodulation. The frequency band signal is demodulated into baseband signal, and bit synchronization determines the sampling decision time of each symbol, that is to distinguish each symbol, so that the receiver can get a series of symbol sequences, which represent certain information. Usually, a number of symbols represent a letter (or symbol, number), while a number of letters form a word, and several words form a sentence. When transmitting data, several code elements will form a code group. The task of group synchronization is to distinguish words, sentences or code groups. For example, if a word is composed of 20 symbols, the timing pulse frequency of the word can be obtained by dividing the symbol bit synchronization pulse frequency FS for 20 times. But the “beginning” and “ending” time of each group, that is, the phase of group synchronization pulse can not be directly obtained from the synchronization pulse, so it needs to be solved by group synchronization.
