Detailed Explanation of Forward Error Correction Technology

As optical communication systems develop toward longer distances, greater capacity, and higher speeds, especially as single-wavelength rates evolve from 40G to 100G and beyond, transmission effects such as chromatic dispersion, nonlinear effects, and polarization mode dispersion in optical fibers will severely impact further increases in transmission rates and distances. To address this, industry experts are continuously researching and developing higher-performing FEC code types to achieve higher net coding gain (NCG) and improved error correction performance to meet the demands of the rapid development of optical communication systems.

1. Meaning and Principle of FEC

FEC (Forward Error Correction) is a method for increasing the reliability of data communications. Optical signals are disrupted during transmission, and the receiving end may mistakenly interpret a "" signal as a "0" signal, or a "0" signal as a "1" signal. Therefore, the FEC function transforms the information code into a code with a certain error correction capability at the transmitting end's channel encoder. The receiving end's channel decoder decodes the received code. If the number of errors generated during transmission is within the error correction capability (non-continuous errors), the decoder locates the error and corrects it, thereby improving signal quality.

2. Two Receive Signal Processing Methods of FEC

FEC Received signal processing methods can be categorized into two main types: hard decision decoding and soft decision decoding. Hard decision decoding is based on traditional error correction code concepts. The demodulator sends the decision result to the decoder, which then uses the algebraic structure of the codeword to correct errors. Soft decision decoding incorporates more channel information than hard decision decoding. The decoder can fully utilize this information through probabilistic decoding, achieving greater coding gain than hard decision decoding.

3. FEC Development History

FEC has gone through three generations in terms of timing and performance.

The first generation FEC uses hard decision block codes, with RS (255, 239) being a typical example. This has been written into the ITU-TG.709 and ITU-TG.975 standards. The codeword overhead is 6.69%, and when the output BER is 1E-13, the net coding gain is about 6dB. The second generation FEC uses hard decision concatenated codes, which combine concatenation, interleaving, and iterative decoding. The codeword overhead is still mainly 6.69%, and when the output BER is 1E-15, the net coding gain is more than 8dB, which can support 10 To meet the long-distance transmission requirements of 100G and 40G systems, third-generation FEC uses soft decision-making, with a codeword overhead of 15%-20%. When the output BER is 1E-15, the net coding gain reaches approximately 1dB, which can support the long-distance transmission requirements of 100G and beyond 100G systems.

4. FEC and 100G Optical Module Applications

FEC is used in high-speed optical modules such as 100G. Enabling this function generally increases the transmission distance of high-speed optical modules compared to without FEC. For example, 100G QSFP28 ZR4 optical modules generally support transmission distances up to 80 km. When FEC is enabled on single-mode fiber, the maximum transmission distance can reach 90 km. However, due to the inevitable delay of data packets during error correction, it is not recommended to enable this function for all high-speed optical modules. For example, when using 100G QSFP28LR4 optical modules, FEC is not recommended. Furthermore, the switch must support FEC. If FEC is enabled on the optical module at end A, it must also be enabled on the optical module at end B, otherwise the interface will not be up.