Combined echo and crosstalk cancellation

#The present invention relates generally to echo and crosstalk cancellation. More particularly, the present invention relates to frequency domain adaptive filters for echo and crosstalk cancellation.

#Communications systems, including 10GBase-T Ethernet systems, may make use of cancellation to improve the accuracy of communications. Noise to be cancelled may include echo and crosstalk.

#Embodiments of the invention provide combined echo and crosstalk cancellation. Frequency domain adaptive filters may be used to remove or reduce the effects of echo and crosstalk for a multi-channel and full-duplex communications system. Data from each transmit channel may be buffered and converted to the frequency domain. The frequency domain data may be multiplied by crosstalk coefficients to obtain a frequency domain correction signal for each channel. Adaptation of the crosstalk coefficients may be based on correlations between the error signals and the data from each of the transmit channels. A single frequency domain transform engine, such as a Fast Fourier Transform engine, may be employed for all calculations to save power and area.

#FIG. 1 is a schematic illustration of a DSP system including an Echo/NEXT block in accordance with an embodiment of the present invention.

#FIG. 2 is a schematic illustration of a frequency domain approach for correlating the error signal to the transmit data in accordance with an embodiment of the present invention.

#FIG. 3 is a schematic illustration of an Echo/NEXT canceller block in accordance with an embodiment of the present invention.

#Echo/NEXT Function

#The Echo/NEXT block fits in the DSP system as shown in FIG. 1. On the transmit data path, the Echo/NEXT block buffers data from the output of the THP, and after the Echo/NEXT correction has been calculated, it outputs the Tx data to the DACs in the AFE. On the receive data path, the Echo/NEXT block uses an error signal from the receiver to adapt the Echo/NEXT, and generates a correction signal that is subtracted from the receive signal.

#The Echo/NEXT uses a frequency domain approach for correlating the error signal to the transmit data. The algorithm is illustrated in FIG. 2. Data from each transmit THP channel is buffered into a echonext frame of samples. Each echonext frame is 512 samples, and echonext frames are not correlated to data or LDPC frames. The data then converted to the frequency domain using an FFT. Data frequency domain transmit data is multiplied by the echo and crosstalk coefficients. The frequency domain correction for each channel is summed from the 4 aggressors. The frequency domain correction is then converted to time domain correction samples with an IFFT.