| Popis: |
Crosstalk in DSL leads to significant performance degradation and large losses in data-rate. Several crosstalk cancellation techniques have been proposed to address this problem, however, in the existing literature the analysis of these approaches is based on SNR calculations and the SNR-gap approximation. Furthermore, for crosstalk cancellation techniques based on decision-feedback, the effect of error propagation is completely ignored. This makes it hard to predict the performance of crosstalk cancellation in real life, and to see if the significant potential gains can actually be realized. To address this problem, this paper uses Monte- Carlo simulation to investigate the performance of the various crosstalk cancellation techniques that have been proposed. The effect of noise-enhancement in zeroforcing crosstalk cancellers and error-propagation in decision-feedback cancellers is examined. The results confirm that a very simple crosstalk cancellation structure can achieve near-optimal performance |
