Reduced-state MLSE for an IM/DD system using PAM modulation

Abstract

The performance of a high-speed intensity-modulation (IM)/direct-detection (DD) transmission system could be limited by the bandwidth of optical transceivers. One popular way to cope with this performance limitation is to utilize the maximum likelihood sequence estimation (MLSE) at the receiver. However, a practical problem of MLSE is its high implementation complexity. Even though the channel impulse response can be truncated by using a two-tap filter before applying the MLSE, it still faces an implementation problem when used for multi-level modulation formats. In this paper, we propose and demonstrate a reduced-state MLSE for band-limited IM/DD transmission systems using M-ary pulse amplitude modulation (PAM-M) formats. We use a conventional Viterbi algorithm to search a reduced-state trellis, which is constructed by using the coarse pre-decision of the signal equalized by a feed-forward equalizer. Thus, the proposed MLSE reduces the implementation complexity significantly. We evaluate the performance of the proposed reduced-state MLSE over 100∼140-Gb/s PAM-4/6/8 transmission systems implemented by using a 1.3-µm directly modulated laser. The results show that the proposed MLSE achieves almost the same performance as the conventional MLSE but reduces the implementation complexity by a factor of 4∼10 when the complexity is assessed by the number of multiplications and additions.

Multi-Wavelength Photonic Neuromorphic Computing for Intra and Inter-Channel Distortion Compensations in WDM Optical Communication Systems

“…The gray curves are the contour lines given by the product of the data rate and the distance. When only considering a single channel, our work can achieve a data rate-distance product of 1200 GBaud•km, outperforming most of the other approaches except [82]. The algorithm used in [82] requires a feedforward equalizer and MLSE.…”

“…When only considering a single channel, our work can achieve a data rate-distance product of 1200 GBaud•km, outperforming most of the other approaches except [82]. The algorithm used in [82] requires a feedforward equalizer and MLSE. Its problem is that in the system with more distinct dispersioninduced pulse broadening, the computational complexity increases enormously.…”

“…We then adapt MLSE structure from [82] to derive the operations number of multiplication N M,mlse and addition N A,mlse . In order to search for the most probable signal sequence using MLSE, the memory length must scale with the dispersion-induced pulse spreading length which leads to the exponentially increased operation number [82], as shown in follows:…”

Machine Learning Applications for Short Reach Optical Communication

“…As a result, more complex approaches such as the maximum likelihood sequence estimation (MLSE) [112] and ML-based equalizers have been studied. ML techniques, especially neural networks, are now considered as a powerful equalization tool to compensate both linear and nonlinear distortions in optical communications.…”

Simplified SVM Equalization Algorithm Based on Single Hyperplane Training Enabled 50Gb/s PAM-4/8 With 10-G Optics in NG-PON System

“…In this way, the further NG-PON needs to be configurated with high transmission speed (above 25Gb/s and 50 Gb/s per-wavelength) while keeping low upgraded cost [5]. To this end, the intensity modulation direct detection (IMDD) technique combined with high data rate based on 10-GHz optoelectronic devices has being considered as one of the most promising solution for large capacity NG-PON due to its simple structure and low cost [6]. Meanwhile, for its high spectral efficiency, the high-order pulse amplitude modulation (PAM- M) formats such as four/eight-level pulse amplitude modulation (PAM-4/8), is also utilized in high-speed NG-PON systems to obtain high data rate [7].…”