Data-defined naïve Bayes (DNB) based decision scheme for the nonlinear mitigation for OAM mode division multiplexed optical fiber communication

Multiplexed optical fiber communication is a technique that allows multiple signals to be transmitted over a single optical fiber. This technique is widely used in modern communication systems because it allows for the efficient utilization of optical fibers, which are capable of transmitting large amounts of data over long distances. However, the capacity of optical fiber systems is approaching limits imposed by nonlinear effects and signal interference. To overcome this challenge, new multiplexing techniques are needed to increase the data-carrying capacity and reduce the signal processing complexity of optical fiber systems. One promising technique is mode division multiplexing (MDM), which is a technique for multiplexing optical signals based on their different modes of propagation within a single optical fiber. In traditional single-mode fiber, only one mode is supported, but in multimode fiber, multiple modes can be propagated. MDM uses the different modes of propagation to transmit multiple signals over a single fiber. It uses different spatial modes of light as independent channels to encode information.

Among various spatial modes, orbital angular momentum (OAM) modes have attracted considerable attention due to their orthogonality, scalability, and robustness. OAM modes are characterized by a helical phase front that carries a well-defined topological charge. OAM mode-division multiplexed optical fiber communication systems have demonstrated terabit-scale data transmission over specially designed fibers that minimize mode coupling and crosstalk. However, OAM mode-division multiplexed optical fiber communication systems still face some challenges, such as the nonlinear effects caused by the opto-electronic devices used in the system. These nonlinear effects can degrade the signal quality and increase the bit error rate (BER) of the system. Therefore, effective nonlinear mitigation techniques are required to improve the performance of OAM mode-division multiplexed optical fiber communication systems. It is important to have a deep understanding of the principles of multiplexed optical fiber communication and the different techniques used to implement it. This includes knowledge of the different types of optical fibers, the characteristics of optical signals, and the equipment used to transmit and receive these signals. It is also important to be familiar with the different standards and protocols used in optical fiber communication, as well as the latest developments in the field.

In a new study published in the peer-reviewed journal Optics Express Professor Ran Gao and co-workers from Beijing institute of technology designed a data-defined naïve Bayes (DNB) algorithm that can effectively mitigate the nonlinear effects caused by the opto-electronic devices in an OAM mode-division multiplexed optical fiber communication system. The study assumed that the prior probability distribution of the transmitted symbols is known and can be used to calculate the posterior probability of each symbol given the received signal. The DNB algorithm then selects the symbol with the highest posterior probability as the decision output.

The research team reported the novel decision scheme for mitigating the nonlinear effects in an optical fiber communication system that uses OAM modes as carriers of information. The scheme is based on a DNB algorithm, which uses the prior probability distribution of the transmitted symbols and the Gaussian distribution of the received signal to calculate the posterior probability of each symbol and select the most likely one as the output. The DNB algorithm is flexible and can work with any modulation format and any number of OAM modes. The study compares the DNB algorithm with the conventional hard decision (HD) and soft decision (SD) schemes based on the Euclidean distance, as well as other machine learning based methods such as support vector machine (SVM) and k-nearest neighbors (KNN). The performance of the DNB algorithm is evaluated by the bit error rate (BER), the constellation diagram, and the eye diagram. The authors also conducted an experiment using a 32GBaud Nyquist QPSK signal as the input data source, an 8-mode OAM multiplexer and demultiplexer to generate and separate eight OAM modes with different topological charges, a 10 km ring core fiber with low inter-mode group crosstalk as the transmission medium, a coherent receiver with a local oscillator to detect the signal in each OAM mode, and a real-time oscilloscope to capture the electrical waveforms and perform offline digital signal processing. The findings showed that the DNB algorithm can achieve a BER reduction of up to 66% compared to the HD and SD schemes for all OAM modes, and has a lower DSP complexity than the SVM and KNN methods, which require more training samples and more computation time. The study demonstrates that the DNB algorithm is an effective and efficient technique for nonlinear mitigation in OAM mode-division multiplexed optical fiber communication systems.

The study concludes that the DNB algorithm is a novel and effective technique for nonlinear mitigation in OAM mode-division multiplexed optical fiber communication systems. Professor Ran Gao and colleagues showed that the DNB algorithm can achieve a significant improvement in the BER performance and the signal quality for all OAM modes compared to conventional and machine learning based methods. They successfully demonstrated that the DNB algorithm has a low computational complexity and does not require complex theoretical models or large training samples.