Optimized MSA DPD method for improving 5G multiband optical fronthaul performance

Abstract

AbstractA modified optimized magnitude‐selective affine (OMSA) model‐based digital predistortion (DPD) is presented that introduces the weighting function into our earlier proposed magnitude‐selective affine (MSA) method with an aim to further reduce the complexity overheads without affecting performance compared to the MSA method. This model utilizes a power‐reliant weighted function rather than the summation of MSA quantities for improving the multiband 5G new radio (NR) analog radio over fiber system performance. The OMSA‐DPD method is tested using 5G NR signals which are transmitted over a 10‐km fiber length. The performance of the OMSA‐DPD method is assessed in comparison to MSA and generalized memory polynomial (GMP) methods in terms of adjacent channel power ratio, error vector magnitude, and complexity. The experimental results show that the OMSA‐DPD method achieves better performance with lower complexity compared to the MSA and GMP models, meeting the 3GPP limits.