Efficient multiplexing using delayed CSI in few-mode fiber links

Abstract

The constant demand for high data rates has resulted in few-mode optical fiber (FMF) based multiplexed systems an attractive solution. While the separability of FMF modes permits effective multiplexing, modal dispersion limits data rates. Conventional signal processing based approaches to compensate for modal dispersion incur significant computational complexity. On the other hand of principal modes (PMs) can mitigate dispersion with very little compensation requirements at the receiver. However, in the presence of high mode-dependent losses (MDL), the PMs’ performance degrades significantly. In this paper, we propose the use of a polar decomposition based scheme, wherein the impairments of the fiber can be viewed as a cascade of a dispersive system and a lossy (MDL) system. Compensating for the MDL using polar decomposition permits the use of PMs with near optimal performance. Further, we discuss effective quantization techniques to track the variation of PMs and propose effective low-rate feedback mechanisms for PM feedback to the transmitter for optimized transmission, with interpolation across WDM channels to reduce the feedback burden. Finally, we also discuss the impact of delay in feedback on achievable rate when the fiber channel undergoes temporal variation. Simulations reveal that the polar decomposition and quantized PM approach yields effective data rates for various fiber conditions, even with temporal channel variations, thereby indicating that this would be a promising solution for low complexity multiplexing based FMF systems.