PLC-to-DSL Interference: Statistical Model and Impact on DSL

TL;DR

This paper characterizes PLC-to-DSL interference using a statistical model, assessing its impact on various DSL technologies and highlighting how loop length and vectoring influence performance degradation.

Contribution

It introduces a novel statistical model for PLC-to-DSL interference based on measured data, enabling accurate impact evaluation on different DSL profiles.

Findings

Interference impact varies with loop length and vectoring.

Average interference impact is small, but worst-case can be significant.

The model accurately predicts both average and worst-case interference effects.

Abstract

Newly available standards for broadband access using Digital Subscriber Lines (DSL) have a high degree of spectrum overlap with home networking technologies using broadband Power Line Communications (BB-PLC) and this overlap leads to Electromagnetic Compatibility issues that may cause performance degradation in DSL systems. This paper studies the characteristics of measured PLC-to-DSL interference and presents novel results on its statistical characterization. The frequency-dependent couplings between power line cables and twisted-pairs are estimated from measurements and a statistical model based on a mixture of two truncated Gaussian distributions is set forth. The proposed statistical model allows the accurate evaluation of the impact of BB-PLC interference on various DSL technologies, in terms of both average and worst-case impacts on data rate. This paper further provides an extensive assessment of the impact of PLC-to-DSL interference at various loop lengths and for multiple profiles of Very-high rate Digital Subscriber Lines (VDSL2), Vectored VDSL2 (V-VDSL2), and G.fast. The results of this paper confirm that the impact of PLC interference varies with loop length and whether vectoring is used or not. Furthermore, the average impact is found to be generally small but worst-case couplings can lead to substantial degradation of DSL.