Project

Modelagem e Mitigação de Interferência em Sistemas G.fast

Goal: O presente projeto pretende avaliar as características do AXT que impactam na correlação espacial da interferência observada no grupo vetorizado e que limitam o desempenho alcançado pelos métodos de mitigação de AXT baseados em predição em situações com transmissão G.fast. Especial atenção é dada a ambos o comportamento a respeito do número de fontes de AXT e ao desbalanceamento resultante dos métodos de mitigação. Este projeto também estenderá o trabalho apresentado em (Gomes et al 2015), motivado pela importância de AXT devido seu impacto em sistemas G.fast.

Updates
0 new
0
Recommendations
0 new
0
Followers
0 new
3
Reads
0 new
60

Project log

Diego De Azevedo Gomes
added a research item
G.fast is one of the most recent DSL system standardized by ITU. It aims to achieve transmission rates of the order of Gbps over short loops, using bandwidth of 106 MHz or 212 MHz. Owing to its large bandwidth, the G.fast is susceptible to interference from a wide range of sources, one of the most damaging being alien crosstalk. This study provides a set of formulations that help the researchers of alien crosstalk mitigation methods to have a better understanding of the behaviour of alien crosstalk, particularly with regard to the impact of the number of the alien crosstalk sources on the performance of the prediction-based alien crosstalk mitigation methods. Additionally, we have carried out a series of simulations and practical experiments to confirm the formulations outlined here.
Diego De Azevedo Gomes
added a research item
Alien crosstalk is one of the major impairments for copper-based transmissions. This letter outlines a method for mitigating alien crosstalk for DSL downstream transmissions impaired by multiple interference sources. The method requires a reference channel, and includes a post processing stage in which induced correlation is applied to prepare the interference at the target channel to be reasonably removed by a prediction based mitigation step. The results show that the proposed method outperforms published alien crosstalk mitigation methods as the number of interference sources increase in G.fast scenarios.
Diego De Azevedo Gomes
added a research item
G.fast is the most recent ITU-T standard for copper-based transmission, which targets at short distances and adopts a bandwidth of 106 MHz extensible to 212 MHz. In several scenarios G.fast systems will be composed by or coexist with uncoordinated lines, which are sources of strong crosstalk. This work focuses on evaluating an alien (out-of-domain) crosstalk mitigation method based on noise prediction that was proposed for predecessors of G.fast. Simulation results show that alien crosstalk mitigation methods can indeed improve the performance of G.fast systems, but under specific conditions. We have contrasted these situations and we provide guidelines about the feasibility of the crosstalk mitigation technique. As support for the simulations, we also present a set of alien crosstalk measurements.
Diego De Azevedo Gomes
added 2 research items
The DSL technologies are expected to face many interferences in conventional deployments, such as RFI. One way to mitigate the destructive effects of RFI is to use the correlation between the RFI observed in differential mode and the RFI observed in common mode. This strategy is considered in many works found in the literature. However, they do not show how to overcome the impairments caused by the signal that leaks from differential to common mode (mode conversion), which causes considerable performance decrease. The work presented here shows how to overcome the conversion mode impairment, through a simple adjust in the current common mode based PLC mitigation methods. Additionally, we show the impact of the mode conversion in the conventional methods, and the SINR improvement achieved by our method.
Diego De Azevedo Gomes
added a project goal
O presente projeto pretende avaliar as características do AXT que impactam na correlação espacial da interferência observada no grupo vetorizado e que limitam o desempenho alcançado pelos métodos de mitigação de AXT baseados em predição em situações com transmissão G.fast. Especial atenção é dada a ambos o comportamento a respeito do número de fontes de AXT e ao desbalanceamento resultante dos métodos de mitigação. Este projeto também estenderá o trabalho apresentado em (Gomes et al 2015), motivado pela importância de AXT devido seu impacto em sistemas G.fast.