The Japanese Break The One Petabit Barrier Per Fibre Pair

The new sport in Japan is breaking long haul fibre optic transmission records. NEC and the Japanese government agency, the National Institute of Communication Technology (NICT), announce new throughput records every 6 to 8 months. It is a major area of research. At the SubOptic 2025 conference NEC representatives did many presentations on their multicore research. NICT has taken a slightly different approach of combining multiple spectrum bands (C+L) with multicore fibre. This is more challenging because there are few C+L amplifier products on the market. In fact, only Subcom has manufactured and deployed such a system on the PLCN cable. Japan by the way has been a hub of innovation in optical networking and other high tech areas like space exploration. However, long term economic prospects are poor due to population decline and the low status of women which means their talents are underutilized. 

What is special about this record breaking effort is the use of a 19 core fibre strand with standard dimensions. In other words, it can be spliced to standard G.652d or G.657a fibre. Network compatibility is always essential in our business given the large base of installed infrastructure. The press releases is vague, but it appears the optical cross talk problem has been tamed as they assert optical losses are low. Just as impressive is the optical amplifier spacing. The total distance traversed was 1808 kilometers with the amplifiers spaced every 82 kilometers or 48 US miles. The bit error rate was not disclosed. Each of the 19 cores used the C+L bands. So we are looking at a single strand doing the work of 38 strands. Each optical amplifier consisted of a C band amplifier and an L band amplifier. The coherent modulation scheme was 16QAM. Each spectrum band produced 180x 100G wavelengths. Some how my math based on these reported figures does not sum to a petabit. I would appreciate anyone who can correct my mistake. 

The challenge for this type of system has always been cross talk, which is light leaking of a core and into other cores. It appears NICT has made significant progress in solving this problem. The other key challenge is designing affordable, compact, and power frugal amplifiers. Affordability comes down to production economies of scale as new technology is always initially expensive. Power consumption for subsea cables is limited by voltage drop and the necessity of keeping the cable light enough for deployment, controlling costs, and easy maintenance. Consequently, I suspect this system is still 5 to 10 years from commercial sale. Not surprisingly, there is no explanation of how the cross talk problem was tamed other than a vague reference to the shape and arrangement of the cores. More details available at  https://www.fibre-systems.com/article/transmission-record-set-using-existing-fibre-cable