No self-introduction
Loading
Profile
Follow
codes (2)
DC balance of 8B10B optical fiber communication
no vote
In the high-speed serial data transmission, the transmitted data is encoded into a self synchronous data stream, that is, the data and clock are combined into a single signal for transmission, so that the receiver can easily and accurately separate the data and clock, and to achieve a satisfactory bit error rate, the key technology is the data coding method in the serial transmission. At present, most high-speed serial standards adopt 8B / 10B coding scheme, such as fibre channel 1, PC I - Express, serial ATA, Ethernet, XAU I, in2 fin IB and and serial connected SCS I, etc. Using 8B / 10B coding technology, the following problems are solved. (1) conversion density: ensure enough signal conversion in data stream. Using 8B / 10B coding method, the number of continuous "1" or continuous "0" in the data stream is not more than 5, so that the PLL at the receiving end can work normally and avoid the data loss caused by clock drift or synchronization loss at the receiving end. (2) DC compensation: in high-speed data transmission lines, differential signal is generally used, and the DC component is required to be as small as possible. 8B / 10B has DC compensation function, that is, DC offset will not appear in the link with the passage of time. (3) Error detection: 8B / 10B code adopts redundancy mode, which encodes 8-bit data and some special characters into 10 bit data according to specific rules. According to these rules, error information in the transmission process can be detected.
baiyecha 0 0
1
The main module code of serial port of pseudo random sequence generator Verilog
no vote
Among them, m sequence is the most famous and simple, and also the most thorough sequence. M sequence is also the basis of studying other sequences. It has balanced sequences and the best autocorrelation property, but the number of cross-correlation families satisfying certain conditions is very small (for m sequences whose order of primitive polynomial is less than or equal to 13, the number of mutually preferred pairs is not more than 6), and the linear complexity is very small. The number of m sequence family sequences is extremely large (when the register series is equal to 6, there are 226 sequences). However, it is difficult to generate, and its cross-correlation characteristics are rarely known, so it is rarely used. The cross-correlation function of gold sequence is 3-valued, the sequence is partially balanced, and has good correlation characteristics. The number of family sequences is relatively large, but it has fatal weakness. The linear complexity of gold sequence is very low, which is only twice that of m sequence with the same length. This restricts the wide application of gold sequence, especially in anti-interference and cryptography. GMW sequence has the advantages of sequence balance, large linear complexity and good autocorrelation (same as m sequence). It is a nonlinear sequence and has more number than m sequence. As a single sequence, GMW Sequences have advantages, but a family of GMW Sequences satisfy certain cross-correlation conditions. Generally, it is not used as address code in multiple access communication. The balance and correlation of cascaded GMW Sequences are the same as GMW Sequences, and the number of families is more than GMW Sequences. Kasami sequences are divided into small set Kasami sequences and large set Kasami sequences. The number of small set Kasami sequence family is large, and the cross-correlation value reaches Welch lower bound. The number of large set Kasami sequence family is very large, and the cross-correlation of small set Kasami sequence family is inferior. They all have the same weakness, the sequence is unbalanced, and the linear complexity is not large (but slightly larger than m, gold sequence). Bent sequence is constructed in the early 1980s. It has the advantages of sequence balance, lower Welch correlation value, large number of family sequences and large linear complexity. It is the best pseudo-random sequence in the 80's and 90's. But bent sequence is difficult to construct, and there is no fast algorithm to meet certain requirements. No sequence is a new type of pseudo-random sequence constructed in the late 1980s. Its outstanding advantages are large linear complexity, and the correlation value can reach Welch lower bound. There are many family sequences, but it has the weakness of sequence imbalance.
baiyecha 0 0
1
No more~
Follows0
Fans0