| dc.description.abstract |
While crosstalk cancellation enables high data rates to be achieved, it is computationally demanding, and development of low complexity cancelers is important. In this paper, we show that the special characteristics of the DSL channel ensure that despite the large number of vectored Twisted Pairs (TPs), large constellation sizes, as well as the large number of tones, Maximum Likelihood receivers are practical, and as computationally efficient as linear receivers whose use has been suggested recently. The proposed receiver uses the Space Alternating Generalized Expectation Maximization (SAGE) iterative receiver, and exploits its ability to converge very rapidly close to the correct solution by initializing the iteration with the output of the Frequency Domain Equalizer (at each TP and tone). Analytical expression for Signal to Noise Ratio (SNR) are derived to demonstrate that the iterative receiver converges in just one iteration. |
en_US |