Dear all, I've heard and even read this term in the context of the performance of coding/modulation schemes. But what does it actually mean? What's so important about determining the pair-wise probability? What's wrong with the triplet-wise probability or any other combination? I'm trying to gain an intuitive understanding of why this is so desirable, but am failing miserably, can someone help with a clear example? Thanks, CW
pair-wise error probability - what's the point?
Started by ●August 4, 2006
Reply by ●August 4, 20062006-08-04
"CW" <prada_white@yahoo.ca> writes:> Dear all, > > I've heard and even read this term in the context of the performance of > coding/modulation schemes. But what does it actually mean? What's so > important about determining the pair-wise probability? What's wrong > with the triplet-wise probability or any other combination? I'm trying > to gain an intuitive understanding of why this is so desirable, but am > failing miserably, can someone help with a clear example? > > Thanks, > > CWHi CW, This term simply refers to the probability of detecting one symbol when another (different) symbol is transmitted. The "pair" is the transmitted symbol and the detected symbol. It is used when determining the union bound since you simply sum over all pairwise error probabilities. -- % Randy Yates % "Rollin' and riding and slippin' and %% Fuquay-Varina, NC % sliding, it's magic." %%% 919-577-9882 % %%%% <yates@ieee.org> % 'Living' Thing', *A New World Record*, ELO http://home.earthlink.net/~yatescr
Reply by ●August 7, 20062006-08-07
Randy Yates wrote: (snip question about pairwise probability)> This term simply refers to the probability of detecting one symbol > when another (different) symbol is transmitted. The "pair" is the > transmitted symbol and the detected symbol.But the detected symbol likely also depends on the previous symbol, and possibly the following symbol, so tripletwise or quadrupletwise could make sense. Statistically, it would take too long to collect enough data, which is probably why it isn't done. -- glen
Reply by ●August 7, 20062006-08-07
glen herrmannsfeldt <gah@ugcs.caltech.edu> writes:> Randy Yates wrote: > > (snip question about pairwise probability) > >> This term simply refers to the probability of detecting one symbol >> when another (different) symbol is transmitted. The "pair" is the >> transmitted symbol and the detected symbol. > > But the detected symbol likely also depends on the previous symbol, > and possibly the following symbol, so tripletwise or quadrupletwise > could make sense.Oh-m'gosh. That would be horribly complex. I think the union bound is derived assuming an AWGN channel - i.e., no memory. -- % Randy Yates % "Though you ride on the wheels of tomorrow, %% Fuquay-Varina, NC % you still wander the fields of your %%% 919-577-9882 % sorrow." %%%% <yates@ieee.org> % '21st Century Man', *Time*, ELO http://home.earthlink.net/~yatescr
Reply by ●August 8, 20062006-08-08
Randy, I sort of understand you, but not really. How does this then tie in with coding schemes (i.e transmission schemes with memory)? I don't see the connection given your last statement. CW Randy Yates wrote:> "CW" <prada_white@yahoo.ca> writes: > > > Dear all, > > > > I've heard and even read this term in the context of the performance of > > coding/modulation schemes. But what does it actually mean? What's so > > important about determining the pair-wise probability? What's wrong > > with the triplet-wise probability or any other combination? I'm trying > > to gain an intuitive understanding of why this is so desirable, but am > > failing miserably, can someone help with a clear example? > > > > Thanks, > > > > CW > > Hi CW, > > This term simply refers to the probability of detecting one symbol > when another (different) symbol is transmitted. The "pair" is the > transmitted symbol and the detected symbol. > > It is used when determining the union bound since you simply sum over > all pairwise error probabilities. > -- > % Randy Yates % "Rollin' and riding and slippin' and > %% Fuquay-Varina, NC % sliding, it's magic." > %%% 919-577-9882 % > %%%% <yates@ieee.org> % 'Living' Thing', *A New World Record*, ELO > http://home.earthlink.net/~yatescr
Reply by ●August 8, 20062006-08-08
"CW" <prada_white@yahoo.ca> writes:> Randy, > > I sort of understand you, but not really. How does this then tie in > with coding schemes (i.e transmission schemes with memory)? I don't > see the connection given your last statement.Hi CW, I haven't seen the term used in that context. Can you point me to an example in which you have? --Randy -- % Randy Yates % "I met someone who looks alot like you, %% Fuquay-Varina, NC % she does the things you do, %%% 919-577-9882 % but she is an IBM." %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO http://home.earthlink.net/~yatescr
Reply by ●August 8, 20062006-08-08
Randy, I wasn't trying to claim that this is where i had first happened upon this term, i'm just curious as to whether it actually means anything when talking about either coding schemes with memory or a channel with memory? Any thoughts? CW Randy Yates wrote:> "CW" <prada_white@yahoo.ca> writes: > > > Randy, > > > > I sort of understand you, but not really. How does this then tie in > > with coding schemes (i.e transmission schemes with memory)? I don't > > see the connection given your last statement. > > Hi CW, > > I haven't seen the term used in that context. Can you point me to > an example in which you have? > > --Randy > > -- > % Randy Yates % "I met someone who looks alot like you, > %% Fuquay-Varina, NC % she does the things you do, > %%% 919-577-9882 % but she is an IBM." > %%%% <yates@ieee.org> % 'Yours Truly, 2095', *Time*, ELO > http://home.earthlink.net/~yatescr
Reply by ●August 8, 20062006-08-08
"CW" <prada_white@yahoo.ca> writes:> Randy, > > I wasn't trying to claim that this is where i had first happened upon > this term, i'm just curious as to whether it actually means anything > when talking about either coding schemes with memory or a channel with > memory? Any thoughts?Not as far as I know, but I'm still junior in digital comm. Perhaps Eric Jacobsen can shed some light? --Randy -- % Randy Yates % "And all that I can do %% Fuquay-Varina, NC % is say I'm sorry, %%% 919-577-9882 % that's the way it goes..." %%%% <yates@ieee.org> % Getting To The Point', *Balance of Power*, ELO http://home.earthlink.net/~yatescr
Reply by ●August 8, 20062006-08-08
Randy Yates wrote:> glen herrmannsfeldt <gah@ugcs.caltech.edu> writes:(snip)>>But the detected symbol likely also depends on the previous symbol, >>and possibly the following symbol, so tripletwise or quadrupletwise >>could make sense.> Oh-m'gosh. That would be horribly complex. I think the union bound > is derived assuming an AWGN channel - i.e., no memory.Actually, gigabit ethernet has adaptive equalization, echo cancellation, and cross-talk cancellation, and baseline wander cancellation. (At least the intel version has those.) That is a little different than symbolwise comparisons, but it isn't easy in either case. -- glen
Reply by ●August 8, 20062006-08-08
On Tue, 08 Aug 2006 16:25:48 -0700, glen herrmannsfeldt <gah@ugcs.caltech.edu> wrote:>Randy Yates wrote: > >> glen herrmannsfeldt <gah@ugcs.caltech.edu> writes: > >(snip) > >>>But the detected symbol likely also depends on the previous symbol, >>>and possibly the following symbol, so tripletwise or quadrupletwise >>>could make sense. > >> Oh-m'gosh. That would be horribly complex. I think the union bound >> is derived assuming an AWGN channel - i.e., no memory. > >Actually, gigabit ethernet has adaptive equalization, echo cancellation, >and cross-talk cancellation, and baseline wander cancellation. >(At least the intel version has those.) > >That is a little different than symbolwise comparisons, but it >isn't easy in either case. > >-- glenBaseline wander cancellation in gigabit ? I believe GBE signalling is DC balanced with enough transitions so it doesn't need it. Maybe you're thinking the BLW cancellation for the 100btx phy in the same chip?
