Machine learning approach on synchronization for FEC enabled channels

Within the problem of modern synchronization systems there are is a task of system synchronization based on structure of information being transmitted over the noisy channel. Here is given a relevant example for the channel with forward error correction codes (FEC enabled channel). The method being redefined here is based on FEC data structure of the transmitted information which includes redundant symbols, and machine learning approach which is using this redundant information to detect synchronization and synchronize the whole system with less spare date being transmitted over the channel, which usually being used for synchronization purposes. Suggested redefined method of synchronization takes more computation efforts and more memory for computation, but on the other hand it set an area for its implementation in terms of complexity of synchronizing machine versus data transmission rate: both being pushed by the level of noise. Implication of machine learning mechanism is giving a great improvement on top of the produced result, providing much better performance for the whole end-to-end system, dramatically decreasing delays of taking decision on synchronization, or extending the range of the system use by improving applicability of the system over higher levels of noise. For wired telecommunication systems this leads to better quality, but for wireless systems it leads to wide range or use or/and better quality. Use of reviewed approach for FEC enabled channels leads to new result for synchronization problems solving. Moreover, proposed approach might provide better decoding level due to soft decisions being forward from synchronization module into decoding module of receiving system.