Finite Word Length Effects on Transmission Rate in Zero Forcing Linear Precoding for Multichannel DSL

TL;DR

This paper analyzes how finite word length in linear zero-forcing precoders affects transmission rates in DSL systems, providing formulas and simulations to guide hardware design for minimal capacity loss.

Contribution

It offers an analytical formula linking precoder word length to transmission rate loss, validated by simulations, aiding hardware design in DSL crosstalk cancellation.

Findings

14-bit precoder representation causes less than 1% capacity loss over 300m lines

Analytical estimates match simulation results accurately

Results are robust against channel estimation errors

Abstract

Crosstalk interference is the limiting factor in transmission over copper lines. Crosstalk cancelation techniques show great potential for enabling the next leap in DSL transmission rates. An important issue when implementing crosstalk cancelation techniques in hardware is the effect of finite world length on performance. In this paper we provide an analysis of the performance of linear zero-forcing precoders, used for crosstalk compensation, in the presence of finite word length errors. We quantify analytically the trade off between precoder word length and transmission rate degradation. More specifically, we prove a simple formula for the transmission rate loss as a function of the number of bits used for precoding, the signal to noise ratio, and the standard line parameters. We demonstrate, through simulations on real lines, the accuracy of our estimates. Moreover, our results are stable in the presence of channel estimation errors. Finally, we show how to use these estimates as a design tool for DSL linear crosstalk precoders. For example, we show that for standard VDSL2 precoded systems, 14 bits representation of the precoder entries results in capacity loss below 1% for lines over 300m.