Joint Detection and Localization of Multipath Interference in IM/DD Systems Using Cyclic Accumulation-Based Signal and Noise Correlation

Abstract

Online detection and localization of multipath interference (MPI) are crucial for the operation and maintenance of intensity modulation direct detection (IM/DD) systems. However, existing schemes are hindered by prohibitive on-chip memory requirements. In this paper, we propose and experimentally demonstrate a joint MPI detection and localization scheme based on signal and noise correlation (SNC) combined with cyclic accumulation (CA). For MPI detection, we demonstrate that the correlation peak height serves as a reliable metric for quantifying MPI severity. For MPI localization, by incorporating the CA operation, the proposed CA-SNC scheme compresses the memory required for correlation from several megabytes to merely 0.5 KB, achieving an over 4000-fold storage reduction. Moreover, we comprehensively discuss the simultaneous localization of multiple reflection paths. By leveraging the knowledge of the link structure alongside a redundant accumulation period, we resolve the peak-to-path matching ambiguity. Experiments on a 26.5625 GBaud PAM4 IM/DD system demonstrate a 3 m localization accuracy using only receiver-side digital signal processing (DSP), validating the scheme as a robust, hardware-efficient solution for in-service MPI detection and localization.