Sensitivity of nonlinear precoding to imperfect channel state information in G.fast. Maes, J., Nuzman, C., & Tsiaflakis, P. In 2016 24th European Signal Processing Conference (EUSIPCO), pages 290-294, Aug, 2016.
Sensitivity of nonlinear precoding to imperfect channel state information in G.fast [pdf]Paper  doi  abstract   bibtex   
Nonlinear Tomlinson-Harashima precoding has been proposed as a near-optimal interference mitigation technique for downstream transmission in G.fast systems, particularly in the transmit spectrum above 106 MHz. We propose an alternative implementation of Tomlinson-Harashima precoding and examine performance of the common and alternative implementations with respect to quantization errors and imperfect channel state information. We show that Tomlinson-Harashima precoding is more sensitive than optimized linear precoding to varying channel state information due to fluctuations in ambient conditions and sudden changes in termination impedance. We also show that Tomlinson-Harashima precoding only outperforms optimized linear precoding when the channel state information is almost perfectly known.
@InProceedings{7760256,
  author = {J. Maes and C. Nuzman and P. Tsiaflakis},
  booktitle = {2016 24th European Signal Processing Conference (EUSIPCO)},
  title = {Sensitivity of nonlinear precoding to imperfect channel state information in G.fast},
  year = {2016},
  pages = {290-294},
  abstract = {Nonlinear Tomlinson-Harashima precoding has been proposed as a near-optimal interference mitigation technique for downstream transmission in G.fast systems, particularly in the transmit spectrum above 106 MHz. We propose an alternative implementation of Tomlinson-Harashima precoding and examine performance of the common and alternative implementations with respect to quantization errors and imperfect channel state information. We show that Tomlinson-Harashima precoding is more sensitive than optimized linear precoding to varying channel state information due to fluctuations in ambient conditions and sudden changes in termination impedance. We also show that Tomlinson-Harashima precoding only outperforms optimized linear precoding when the channel state information is almost perfectly known.},
  keywords = {channel capacity;channel coding;digital subscriber lines;nonlinear codes;precoding;quantisation (signal);quantization errors;G.fast systems;downstream transmission;near-optimal interference mitigation technique;nonlinear Tomlinson-Harashima precoding;imperfect channel state information;Precoding;Quantization (signal);Channel state information;Crosstalk;Matrix decomposition;G.fast;nonlinear precoding;Tomlinson-Harashima Precoding;vectoring;sudden termination change},
  doi = {10.1109/EUSIPCO.2016.7760256},
  issn = {2076-1465},
  month = {Aug},
  url = {https://www.eurasip.org/proceedings/eusipco/eusipco2016/papers/1570256207.pdf},
}
Downloads: 0