USRE45794EActiveUtility

Crosstalk cancellation using sliding filters

50
Assignee: MARVELL INT LTDPriority: Sep 26, 2007Filed: May 24, 2013Granted: Nov 3, 2015
Est. expirySep 26, 2027(~1.2 yrs left)· nominal 20-yr term from priority
H04B 3/32
50
PatentIndex Score
0
Cited by
215
References
25
Claims

Abstract

A crosstalk cancellation system and method is disclosed for use in a multi-channel communication system. Crosstalk which couples between channels is cancelled through use of an in-line FFE filter and in-line delay. A cross-connect system associated with each channel includes a cross-connect delay and cross-connect filter. The cross-connect system generates a cancellation signal for each of the channels, which is routed into a junction. The junction subtracts the cancellation signal from the received signal, which has also been delayed and filtered, to remove unwanted cross-talk. During training, cancellation magnitude is monitored at various delay offsets to determine which offset and corresponding filter coefficients, for each delay, maximizes cancellation. The filters are set with filter coefficients that maximize cancellation. The cross-connect delay with the maximum offset is set to zero and its calculated offset amount is established as the in-line delay offset. The other cross-connect delay offsets are adjusted accordingly.

Claims

exact text as granted — not AI-modified
What is claimed is: 
     
       1. A method for training a crosstalk cancellation system that comprises one or more in-line filters, one or more in-line delays, one or more cross-connect filters, and one or more cross-connect delays, the method comprising, for each channel the method comprising:
 providing a crosstalk cancellation system comprising one or more in-line filters, one or more in-line delays, one or more cross-connect filters, one or more cross-connect delays and a memory; 
 setting one or more in-line delays and one or more cross-connect delays to zero delay; 
 training the one or more in-line filters; 
 training the one or more cross-connect filters to generate cross-connect filter coefficients; 
 storestoring the cross-connect filter coefficients; 
 continuecontinuing to adapt the one or more in-line filters and store resulting in-line filter coefficients with the cross-connect filter coefficient as filter coefficient sets; 
 establishing an offset in one or more in-line delays; 
 retraining the one or more in-line filters and the one or more cross-talk filters with the established offset to establish new coefficient set; 
 storing the new coefficient set; 
 establishing additional offsets and for each additional offset: 
 retraining the one or more in-line filters and the one or more cross-talk filters to create a filter coefficient set associated with each offset; 
 storing the additional filter coefficients sets associated with each offset; 
 analyzeanalyzing the filter coefficient sets to determine which offset maximizes cancellation to thereby identify a maximum cancellation offset; 
 establishestablishing one or more in-line filters with coefficient set which corresponds to maximum cancellation offset; 
 establishestablishing one or more in-line delays with offset which corresponds to maximum cancellation offset for victim-interferer pair that needs the largest delay to maximize its cancellation; 
 establishestablishing one or more cross-connect filters with a coefficient set which corresponds to maximum cancellation offset; and  
 establishestablishing one or more cross-connect delays with offset which corresponds to maximum cancellation offset. 
 
     
     
       2. The method of  claim 1 , further comprising:
 analyzing the offset for the one or more cross-connect delays to determine a cross-connect delay that represents the relative delay needed between victim and interferer pair to maximize cancellation; 
 establishing the one or more in-line delays to have a delay amount which maximizes cancellation; and  
 adjust cross-connect delays to maintain same amount of delay at output by taking into account newly establish in-line delay amount. 
 
     
     
       3. The method of  claim 1 , wherein the one or more in-line filters comprises one or more in-line feed forward equalizer type filters. 
     
     
       4. The method of  claim 1 , wherein the method is performed in a four channel communication system. 
     
     
       5. The method of  claim 1 , wherein during training one or more coefficient set is stored in a memory of the crosstalk cancellation system. 
     
     
       6. A crosstalk cancellation system comprising:
 a plurality of receivers corresponding to a plurality of respective communication channels, wherein each receiver in the plurality of receivers includes
 an in-line filter, 
 an in-line delay, and 
 a cross-connect system that includes, for each other receiver in the plurality of receivers, (i) a cross-connect filter in a path coupled to the other receiver and (ii) a cross-connect delay in the path coupled to the other receiver; and 
   a processor configured to, for each receiver in the plurality of receivers, establish settings for (i) the in-line filter, (ii) the in-line delay, (iii) each of the cross-connect filters, and (iv) each of the cross-connect delays, at least in part by
 iteratively training, with a different one of a plurality of in-line delay values being applied to the in-line delay during each iteration, (i) the in-line filter and (ii) at least one of the cross-connect filters to generate a plurality of sets of filter coefficients, 
 identifying (i) a first in-line delay value of the plurality of in-line delay values, and (ii) a first set of filter coefficients, of the plurality of sets of filter coefficients, that maximize crosstalk cancellation, 
 setting the in-line filter based at least in part on the first set of filter coefficients, and 
 setting the in-line delay based at least in part on the first in-line delay value.  
   
     
     
       7. The crosstalk cancellation system of claim 6, wherein the processor is configured to establish the settings further by:
 determining a plurality of cross-connect delay values at least in part by determining, for each other receiver in the plurality of receivers, a cross-connect delay value that maximizes cancellation with respect to the other receiver; and   identifying a maximum cross-connect delay value of the plurality of cross-connect delay values,   wherein setting the in-line delay is based at least in part on (i) the first in-line delay value and (ii) the maximum cross-connect delay value.    
     
     
       8. The crosstalk cancellation system of claim 6, wherein the in-line filter is a feed-forward equalizer (FFE) filter.  
     
     
       9. The crosstalk cancellation system of claim 16, wherein the processor is configured to establish the settings further by:
 prior to iteratively training the in-line filter and the at least one of the cross-connect filters, setting the in-line delay and each of the cross-connect delays to zero delay.    
     
     
       10. The crosstalk cancellation system of claim 6, wherein the processor is configured to iteratively train the in-line filter at least by iteratively training the in-line filter to maximize a signal-to-noise ratio.  
     
     
       11. The crosstalk cancellation system of claim 6, wherein the processor is configured to iteratively train the at least one of the cross-connect filters at least by iteratively training the at least one of the cross-connect filters to maximize crosstalk cancellation.  
     
     
       12. The crosstalk cancellation system of claim 6, further comprising a memory, wherein the processor is configured to establish the settings further by:
 storing a plurality of data sets in the memory, wherein each of the plurality of data sets includes (i) a respective in-line delay value of the plurality of in-line delay values and (ii) a respective set of filter coefficients, of the plurality of sets of filter coefficients, that corresponds to the respective in-line delay value of the plurality of in-line delay values.    
     
     
       13. The crosstalk cancellation system of claim 22, wherein the processor is configured to identify the first in-line delay value and the first set of filter coefficients at least by determining which data set, of the plurality of data sets, maximizes crosstalk cancellation.  
     
     
       14. The crosstalk cancellation system of claim 6, wherein the processor is configured to iteratively train the in-line filter and the at least one of the cross-connect filters at least by utilizing least mean square (LMS) training.  
     
     
       15. A method for training a crosstalk cancellation system, wherein the crosstalk cancellation system comprises a plurality of receivers corresponding to a plurality of respective communication channels, wherein each receiver in the plurality of receivers includes (i) an in-line filter, (ii) an in-line delay, and (iii) a cross-connect system that includes, for each other receiver in the plurality of receivers, a cross-connect filter in a path coupled to the other receiver and a cross-connect delay in the path coupled to the other receiver, the method comprising, for each receiver in the plurality of receivers:
 establishing settings for (i) the in-line filter, (ii) the in-line delay, (iii) each of the cross-connect filters, and (iv) each of the cross-connect delays, wherein establishing the settings includes
 iteratively training, with a different one of a plurality of in-line delay values being applied to the in-line delay during each iteration, (i) the in-line filter and (ii) at least one of the cross-connect filters to generate a plurality of sets of filter coefficients, 
 identifying (i) a first in-line delay value of the plurality of in-line delay values, and (ii) a first set of filter coefficients, of the plurality of sets of filter coefficients, that maximize crosstalk cancellation, 
 setting the in-line filter based at least in part on the first set of filter coefficients; and 
 setting the in-line delay based at least in part on the first in-line delay value.  
   
     
     
       16. The method of claim 15, wherein establishing the settings further includes:
 determining a plurality of cross-connect delay values at least in part by determining, for each other receiver in the plurality of receivers, a cross-connect delay value that maximizes cancellation with respect to the other receiver; and   identifying a maximum cross-connect delay value of the plurality of cross-connect delay values,   wherein setting the in-line delay is based at least in part on (i) the first in-line delay value and (ii) the maximum cross-connect delay value.    
     
     
       17. The method of claim 15, wherein establishing the settings further includes:
 prior to iteratively training the in-line filter and the at least one of the cross-connect filters, setting the in-line delay and each of the cross-connect delays to zero delay.    
     
     
       18. The method of claim 15, wherein iteratively training the in-line filter includes iteratively training the in-line filter to maximize a signal-to-noise ratio.  
     
     
       19. The method of claim 15, wherein iteratively training the at least one of the cross-connect filters includes iteratively training the at least one of the cross-connect filters to maximize crosstalk cancellation.  
     
     
       20. The method of claim 15, wherein iteratively training the in-line filter and the at least one of the cross-connect filters includes utilizing least mean square (LMS) training.  
     
     
       21. A tangible, non-transitory computer-readable medium storing instructions that, when executed by a processor in a crosstalk cancellation system that comprises a plurality of receivers corresponding to a plurality of respective communication channels, wherein each receiver in the plurality of receivers includes (i) an in-line filter, (ii) an in-line delay, and (iii) a cross-connect system that includes, for each other receiver in the plurality of receivers, a cross-connect filter in a path coupled to the other receiver and a cross-connect delay in the path coupled to the other receiver, cause the processor to, for each receiver in the plurality of receivers:
 establish settings for (i) the in-line filter, (ii) the in-line delay, (iii) each of the cross-connect filters, and (iv) each of the cross-connect delays, wherein the instructions cause the processor to establish the settings at least by
 identifying (i) a first in-line delay value of the plurality of in-line delay values, and (ii) a first set of filter coefficients, of the plurality of sets of filter coefficients, that maximize crosstalk cancellation, 
 setting the in-line filter based at least in part on the first set of filter coefficients, and 
 setting the in-line delay based at least in part on the first in-line delay value.  
   
     
     
       22. The tangible, non-transitory computer-readable medium of claim 21, wherein the instructions further cause the processor to establish the settings by:
 determining a plurality of cross-connect delay values at least in part by determining, for each other receiver in the plurality of receivers, a cross-connect delay value that maximizes cancellation with respect to the other receiver; and   identifying a maximum cross-connect delay value of the plurality of cross-connect delay values,   wherein the instructions cause the processor setting the in-line delay is based at least in part on (i) the first in-line delay value and (ii) the maximum cross-connect delay value.    
     
     
       23. The tangible, non-transitory computer-readable medium of claim 21, wherein the instructions further cause the processor to establish the settings by:
 prior to iteratively training the in-line filter and the at least one of the cross-connect filters, setting the in-line delay and each of the cross-connect delays to zero delay.    
     
     
       24. The tangible, non-transitory computer-readable medium of claim 21, wherein the instructions cause the processor to iteratively train the in-line filter at least by training the in-line filter to maximize a signal-to-noise ratio.  
     
     
       25. The tangible, non-transitory computer-readable medium of claim 21, wherein the instructions cause the processor to iteratively train the at least one of the cross-connect filters at least by iteratively training the at least one of the cross-connect filters to maximize crosstalk cancellation.

Cited by (0)

No later patents cite this yet.

References (0)

No backward citations on record.