![\"the](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2022\/07\/\u56fe\u72471-300x113.png\")
<\/p>\nWhen we read an article, if there is a typographical error, say two words in the wrong order, there will be no difficulty for us to understand the original text. But if there are too many spelling mistakes, it will be difficult for readers to understand the article. At this time, information cannot be obtained correctly and efficiently.<\/p>\n
FEC (Forward Error Correction) works on a similar principle. Signals are encoded as \u201c0\u201d and \u201c1\u201d for transmission, with inevitable degradation and error codes. When this level of error is within the range of FEC’s error correction capability, the system can achieve error-free reception, and thus without the need for re-transmission.<\/p>\n
Hamming Code, probably the first form of FEC, was first invented by Richard Hamming in 1950. While working at Bell Labs, he was annoyed by the frequent errors in punch cards (which were used to record and transmit data at the time), so he devised a coding method to identify and correct errors, thus avoiding the need to copy and resend cards.<\/p>\n
The two essential development directions of fiber optic communication are to increase the transmission rate and to extend the transmission distance. As the transmission rate increases, more factors limit the transmission distance during signal transmission. Chromatic dispersion, nonlinear effects, polarization mode dispersion, and other factors affect the simultaneous enhancement of the two directions. Industry experts have proposed the Forward Error Correction function to reduce the impact of these adverse factors.<\/p>\n
In optical transmission systems, the central role of FEC is to reduce the tolerance of the OSNR of the system. If we compare the optical transmission system to the process of reading, FEC improves the readers\u2019 comprehension, enriches their discriminating experience, and to some extent allows more errors in the article.<\/p>\n
<\/p>\n
Figure1: the schematic diagram of the FEC function<\/span><\/strong><\/p>\n Therefore, we define FEC (Forward Error Correction) as an ability that ensures that the communication system can still achieve error-free transmission under the influence of noise and other impairments. Essentially, FEC is a process of encoding and decoding, and the result of the algorithm is sent as additional information along with the data from the transmitter. By repeating the same algorithm at the far end, the receiver can detect single-bit-level errors and correct them (correctable errors) without re-transmitting the data.<\/p>\n To measure this capability, it is necessary to focus on four\u00a0quantities\u00a0of FEC: pre-correction BER tolerance, coding gain (CG), overhead (OH), and net coding gain (NCG). Let’s take a look at the definition of NCG coding gain: it defines the difference between the Q value corresponding to a certain level of BER (for example, 1 \u00d7 10-15) and the Q (dB) value corresponding to the pre-correction BER tolerance.<\/p>\n Figure 2:\u00a0the coding gain between a certain level of BEC with FEC and without FEC<\/span><\/strong><\/p>\n NCG can be compared to the ability difference to correct and receive the right information between a novice and an expert. Generally speaking, there are two types of FEC technologies: in-band FEC and out-of-band FEC.<\/p>\n In the DWDM\/OTN system, we mainly use out-of-band FEC technology. In G.709, Reed Solomon FEC (RS-FEC) is defined for the OTN system, which is located in the FEC overhead of the OTUk layer, and its location is shown in the following figure.<\/p>\n Figure3\uff1athe location of the RS-FEC\u00a0in G.709<\/span><\/strong><\/p>\n At present, FEC has developed for many generations.<\/p>\n Figure4: The three generations of FEC<\/span><\/strong><\/p>\n In the first and second-generation FEC technologies, decoding usually uses only the algebraic structure of the code. The binary sequence is supplied to the decoder by the demodulator, that is, the demodulator only performs 0, 1 decision on the received sequence. This decoding method is called Hard-Decision (HD-FEC). Different kinds of hard decision FEC are compared as follows\uff1a<\/p>\n\n Table one: Comparisons of three different kinds of hard decision FEC<\/span><\/strong><\/p>\n The Soft-Decision used\u00a0in the third generation of FEC (SD-FEC) is a probabilistic decoding method. It performs multi-bit quantization on the sampled voltage output by the demodulator, and then sends it to the decoder to decode the algebraic structure of the code.<\/p>\n Figure5: Schematic diagram of soft decision technology<\/span><\/strong><\/p>\n As shown in the figure above, the hard decision only uses one threshold to quantize one bit, while the soft decision uses multiple thresholds to quantize the recovered symbols, getting one-bit information plus several-bit probability (confidence) information. It is equivalent to adding Maybe between YES and NO. With the same overhead ratio, SD-FEC\u2019s NCG gain is 1-1.5dB higher than that of hard-decision HFEC.<\/p>\n\n![\"the](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2022\/07\/\u56fe\u72472-300x172.png\")
<\/p>\n\n
![\"the](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2022\/07\/\u56fe\u72473-300x123.png\")
<\/p>\n\n
![\"The](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2022\/07\/\u56fe\u72474-300x133.png\")
<\/p>\n\n\n
\n\t \nCoding<\/th> Encoding Algorithm<\/th> Coding Gain<\/th> Line Speed<\/th> Standard<\/th>\n<\/tr>\n<\/thead>\n \n\t Out-of-band FEC<\/td> RS(255,239)<\/td> 5~7dB <\/td> 10.7Gbps<\/td> G.709<\/td>\n<\/tr>\n \n\t Enhanced-FEC<\/td> RS(255,238)
\nRS(245,210)<\/td>7~9dB<\/td> 12.5Gbps<\/td> No<\/td>\n<\/tr>\n \n\t Advanced-FEC<\/td> RS(255,238)
\nBCH(900,860)
\nBCH(500,491)<\/td>7~9dB<\/td> 10.7Gbps.<\/td> G.709<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n ![\"Schematic](\"https:\/\/www.fibermall.com\/blog\/wp-content\/uploads\/2022\/07\/\u56fe\u72475-300x122.png\")
<\/p>\n