Hi There is a old topic called "LMS implementatiom for echo cancelation ad co-channel interference supression" which says we can use LMS to suppress co-channel interference. I agree adaptive Equalizer helps against co-channel interference. But I don't agree regarding NTSC interference in ATSC. I had done this experiment and couldn't meet the performance required. ATSC standard suggests to use comb filter along with 16 state trellis decoder to cancel interference. It degrades the system performance by >= 3dB. I had tried 1 tap LMS to cancel tone but didn't get required performance. I also tried notching out frequencies near NTSC but it also fails to meet required performance of 2.5D/U. Any effort to suppress NTSC before equalizer doesn't work as Equalizer undo it. It becomes a chicken egg problem and in real time scenario we can't disable equalizer and with equalizer on we can't suppress NTSC tone. -Mishra
LMS for co-channel Interference supression
Started by ●December 21, 2010
Reply by ●December 21, 20102010-12-21
On 12/21/2010 08:01 AM, Mishra wrote:> [...] > I had tried 1 tap LMS to cancel tone but didn't get required performance.Yeah, you're not gonna cancel a tone with a gain control. -- Randy Yates % "My Shangri-la has gone away, fading like Digital Signal Labs % the Beatles on 'Hey Jude'" yates@digitalsignallabs.com % http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO
Reply by ●December 21, 20102010-12-21
On Dec 21, 1:50�pm, Randy Yates <ya...@ieee.org> wrote:> On 12/21/2010 08:01 AM, Mishra wrote: > > > [...] > > I had tried 1 tap LMS to cancel tone but didn't get required performance. > > Yeah, you're not gonna cancel a tone with a gain control. > -- > Randy Yates � � � � � � � � � � �% "My Shangri-la has gone away, fading like > Digital Signal Labs � � � � � � �% �the Beatles on 'Hey Jude'" > ya...@digitalsignallabs.com � � �%http://www.digitalsignallabs.com% 'Shangri-La', *A New World Record*, ELONo one said the one tap couldn't be offset, so it could be an adaptive gain and a fixed delay (better have a real convenient, real stable frequency though). The one tap in the LMS could be complex to get adjustable gain and phase correction for a single frequency. Widrow describes this in his classic paper, applied to power line frequency with harmonics (used 2 real taps to implement ). So Mishra, what are you actually doing with the tone? Dirk
Reply by ●December 22, 20102010-12-22
>On Dec 21, 1:50=A0pm, Randy Yates <ya...@ieee.org> wrote: >> On 12/21/2010 08:01 AM, Mishra wrote: >> >> > [...] >> > I had tried 1 tap LMS to cancel tone but didn't get requiredperformanc=>e. >> >> Yeah, you're not gonna cancel a tone with a gain control. >> -- >> Randy Yates =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0% "My Shangri-lah=>as gone away, fading like >> Digital Signal Labs =A0 =A0 =A0 =A0 =A0 =A0 =A0% =A0the Beatles on 'HeyJ=>ude'" >> ya...@digitalsignallabs.com =A0 =A0=A0%http://www.digitalsignallabs.com%=> 'Shangri-La', *A New World Record*, ELO > >No one said the one tap couldn't be offset, so it could be an adaptive >gain and a fixed delay (better have a real convenient, real stable >frequency though). > >The one tap in the LMS could be complex to get adjustable gain and >phase correction for a single frequency. Widrow describes this in his >classic paper, applied to power line frequency with harmonics (used 2 >real taps to implement ). > >So Mishra, what are you actually doing with the tone? > >Dirk >Hi Dirk I agree with you. I do cancel visual carrier of NTSC interference to ATSC using one tap(Complex Tap). NTSC interference into ATSC is a classical problem to solve. As I mentioned standard suggested some solution but that has limitation and also degrades the performance. I was trying to solve this problem and also at-least meet the requirement using some alternative method. Hence gave a try using 1 Complex tap LMS but couldn't meet the performance. Problem comes while equalizing signal. It fills back the frequency component removed by 1 tap LMS/Notch filter along with frequency components of NTSC. -Mishra
Reply by ●December 24, 20102010-12-24
On 12/21/2010 11:42 PM, Mishra wrote:> [...] > I agree with you. I do cancel visual carrier of NTSC interference to ATSC > using one tap(Complex Tap). > > NTSC interference into ATSC is a classical problem to solve. As I mentioned > standard suggested some solution but that has limitation and also degrades > the performance. > > > I was trying to solve this problem and also at-least meet the requirement > using some alternative method. Hence gave a try using 1 Complex tap LMS but > couldn't meet the performance. > > Problem comes while equalizing signal. It fills back the frequency > component removed by 1 tap LMS/Notch filter along with frequency components > of NTSC.Aren't most NTSC transmitters gone now? Are you trying to solve a problem that no longer exists? -- Randy Yates % "My Shangri-la has gone away, fading like Digital Signal Labs % the Beatles on 'Hey Jude'" yates@digitalsignallabs.com % http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO
Reply by ●December 24, 20102010-12-24
>On 12/21/2010 11:42 PM, Mishra wrote: >> [...] >> I agree with you. I do cancel visual carrier of NTSC interference toATSC>> using one tap(Complex Tap). >> >> NTSC interference into ATSC is a classical problem to solve. As Imentioned>> standard suggested some solution but that has limitation and alsodegrades>> the performance. >> >> >> I was trying to solve this problem and also at-least meet therequirement>> using some alternative method. Hence gave a try using 1 Complex tap LMSbut>> couldn't meet the performance. >> >> Problem comes while equalizing signal. It fills back the frequency >> component removed by 1 tap LMS/Notch filter along with frequencycomponents>> of NTSC. > >Aren't most NTSC transmitters gone now? Are you trying to solve a >problem that no longer exists? >-- >Randy Yates % "My Shangri-la has gone away, fadinglike>Digital Signal Labs % the Beatles on 'Hey Jude'" >yates@digitalsignallabs.com % >http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*,ELO>Hi Randy Thanks for your reply. But it still co-exists and needs some mechanism to cancel the effects. It will take some time to stop analog transmission. Even if problem will not exist, for an engineer its an interesting problem to solve. I am always interested to find alternative ideas other then ATSC suggested. Keep updating if you get some ideas. -Mishra
Reply by ●December 24, 20102010-12-24
On Fri, 24 Dec 2010 00:15:30 -0500, Randy Yates <yates@ieee.org> wrote:>On 12/21/2010 11:42 PM, Mishra wrote: >> [...] >> I agree with you. I do cancel visual carrier of NTSC interference to ATSC >> using one tap(Complex Tap). >> >> NTSC interference into ATSC is a classical problem to solve. As I mentioned >> standard suggested some solution but that has limitation and also degrades >> the performance. >> >> >> I was trying to solve this problem and also at-least meet the requirement >> using some alternative method. Hence gave a try using 1 Complex tap LMS but >> couldn't meet the performance. >> >> Problem comes while equalizing signal. It fills back the frequency >> component removed by 1 tap LMS/Notch filter along with frequency components >> of NTSC. > >Aren't most NTSC transmitters gone now? Are you trying to solve a >problem that no longer exists?Many translaters are still NTSC analog. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.abineau.com
Reply by ●December 24, 20102010-12-24
On 12/24/2010 12:15 PM, Eric Jacobsen wrote:> On Fri, 24 Dec 2010 00:15:30 -0500, Randy Yates<yates@ieee.org> > wrote: > >> On 12/21/2010 11:42 PM, Mishra wrote: >>> [...] >>> I agree with you. I do cancel visual carrier of NTSC interference to ATSC >>> using one tap(Complex Tap). >>> >>> NTSC interference into ATSC is a classical problem to solve. As I mentioned >>> standard suggested some solution but that has limitation and also degrades >>> the performance. >>> >>> >>> I was trying to solve this problem and also at-least meet the requirement >>> using some alternative method. Hence gave a try using 1 Complex tap LMS but >>> couldn't meet the performance. >>> >>> Problem comes while equalizing signal. It fills back the frequency >>> component removed by 1 tap LMS/Notch filter along with frequency components >>> of NTSC. >> >> Aren't most NTSC transmitters gone now? Are you trying to solve a >> problem that no longer exists? > > Many translaters are still NTSC analog.Interesting. Out of the total US OTA TV customer channels, what percentage would you say are still being served by NTSC? -- Randy Yates % "My Shangri-la has gone away, fading like Digital Signal Labs % the Beatles on 'Hey Jude'" yates@digitalsignallabs.com % http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO