Hybrid IRS and DFE Detection Scheme for MDL-Impaired Transmission With TAST Coding

Abstract

Linear threaded algebraic space-time (TAST) codes have shown potential advantage in mitigating the mode-dependent loss in the few-mode-fiber-based mode-division multiplexing (MDM) system. However, the maximum-likelihood (ML) detector employed at the receiver of the MDM system greatly suffers from exponential computational complexity. In this work, we first examine the necessity of space-time coding in the MDM system. On this basis, a new hybrid detection, which integrates the improved reduced-search (IRS) method and the decision feedback equalizer (DFE), is proposed for the TAST-assisted MDM system, in which DFE is employed to detect the more reliable tributaries to obtain a near-ML solution, while IRS is responsible for finding the best solution in the remaining set with a reduced number of candidates. Simulation results show that the proposed detection method can achieve near-optimal solutions and has a low computational complexity. Moreover, it has the attractive advantage of flexible adjustment in the tradeoff between system performance and computational complexity.