Hello, everybody. I read that for MSK-modulated signals with diff coding non-coherent demodualtor is bad because one bit error forces all bits after it to be wrong detected. And coherent detector in contrast is free of this effect. If I use noncoherent detector I map +pi/2 to 0 and -pi/2 to 1; if I use coherent detector I have QPSK-like constellation and track phase shifts also. What is different in detectors in part of symbols mapping, and why I can't to use non-coherent this with different coded symbols? Thank you. Pavel Schukin.
Is a coherent demodulator preferred for GMSK with differential coding
Started by ●September 24, 2007
Reply by ●September 24, 20072007-09-24
Pavel Schukin wrote:> Hello, everybody. > I read that for MSK-modulated signals with diff coding non-coherent > demodualtor is bad because one bit error forces all bits after it to be=> wrong detected. And coherent detector in contrast is free of this effec=t. =CA=C3/A=CC> If I use noncoherent detector I map +pi/2 to 0 and -pi/2 to 1; if I use=> coherent detector I have QPSK-like constellation and track phase shifts=> also. What is different in detectors in part of symbols mapping, and w=hy I> can't to use non-coherent this with different coded symbols?=20=D1=EB=E8=F8=EA=E0=EC =EC=ED=EE=E3=E0 =E1=F3=EA=E0=F4: =ED=E5=F7=EE =ED=E8= =EF=EE=ED=E8=EB.> Thank you. > Pavel Schukin.=C2=FB=EF=E5=E9 =E8=E0=E4=F3. VLV
Reply by ●September 24, 20072007-09-24
Pavel Schukin wrote:> Hello, everybody. > I read that for MSK-modulated signals with diff coding non-coherent > demodualtor is bad because one bit error forces all bits after it to be > wrong detected. And coherent detector in contrast is free of this effect. > If I use noncoherent detector I map +pi/2 to 0 and -pi/2 to 1; if I use > coherent detector I have QPSK-like constellation and track phase shifts > also. What is different in detectors in part of symbols mapping, and why I > can't to use non-coherent this with different coded symbols? > Thank you. > Pavel Schukin.The phenomenon you refer to is sometimes called "error extension". A common way to noncoherently detect MSK is to use a pair of sinx/x shaped bandpass filters centered at the mark and space frequencies -- in other words an ordinary FSK demodulator. The detection statistic is the difference between the envelopes of the filter outputs. Differential encoding is not needed in this case because there is no ambiguity in the frequencies. Another approach is to use an FM discriminator followed by a one bit integrate and dump. Again, differential encoding is not needed. A coherent detector will typically produce a phase ambiguity that can be resolved by differentially encoding and decoding, and in this case error extension will be observed. Another way to resolve phase ambiguity without differential encoding is to search for various permutations of a unique word in the transmitted data. John
Reply by ●September 24, 20072007-09-24
Vladimir Vassilevsky wrote:> > > Pavel Schukin wrote: > >> Hello, everybody. >> I read that for MSK-modulated signals with diff coding non-coherent >> demodualtor is bad because one bit error forces all bits after it to be >> wrong detected. And coherent detector in contrast is free of this effect. > > ��/A� > >> If I use noncoherent detector I map +pi/2 to 0 and -pi/2 to 1; if I use >> coherent detector I have QPSK-like constellation and track phase shifts >> also. What is different in detectors in part of symbols mapping, and >> why I >> can't to use non-coherent this with different coded symbols? > > ������� ����� �����: ���� �������. > >> Thank you. >> Pavel Schukin. > > ����� ����. > > > VLV >Google Translate didn't help much with that one: Slishkam much bukaf: necho niponil Take iadu
Reply by ●September 24, 20072007-09-24
Vladimir Vassilevsky wrote:> > > Pavel Schukin wrote: > >> Hello, everybody. >> I read that for MSK-modulated signals with diff coding non-coherent >> demodualtor is bad because one bit error forces all bits after it to be >> wrong detected. And coherent detector in contrast is free of this effect. > > ��/A� > >> If I use noncoherent detector I map +pi/2 to 0 and -pi/2 to 1; if I use >> coherent detector I have QPSK-like constellation and track phase shifts >> also. What is different in detectors in part of symbols mapping, and >> why I can't to use non-coherent this with different coded symbols? > > ������� ����� �����: ���� �������. > >> Thank you. >> Pavel Schukin. > > ����� ����. > > > VLVand I couldn't read it. Jerry -- Engineering is the art of making what you want from things you can get.
Reply by ●September 24, 20072007-09-24
Jerry Avins wrote:>=20 > and I couldn't read it.That's the Russian inet slung. Here is a literal translation: =D0=9A=D0=93/AM (KG/AM): "Concept is shit/author is dull fuck". =D0=A1=D0=BB=D0=B8=D1=88=D0=BA=D0=B0=D0=BC =D0=BC=D0=BD=D0=BE=D0=B3=D0=B0= =D0=B1=D1=83=D0=BA=D0=B0=D1=84: =D0=BD=D0=B5=D1=87=D0=BE =D0=BD=D0=B8=D0= =BF=D0=BE=D0=BD=D0=B8=D0=BB. Too many characters; understood nothing. =D0=92=D1=8B=D0=BF=D0=B5=D0=B9 =D0=B8=D0=B0=D0=B4=D1=83. Drink a poison. What should be the proper English translation? VLV
Reply by ●September 24, 20072007-09-24
On Mon, 24 Sep 2007 19:29:50 -0500, Vladimir Vassilevsky wrote:> Jerry Avins wrote: >> >> and I couldn't read it. > > That's the Russian inet slung. > Here is a literal translation: > > КГ/AM (KG/AM): > "Concept is shit/author is dull fuck". > > Слишкам многа букаф: нечо нипонил. > Too many characters; understood nothing. > > Выпей иаду. > Drink a poison. > > What should be the proper English translation? > > VLVOn USENET, generally, blessed silence. Yelling at the naeve diminishes you and discourages them from asking the questions that will get them less naeve, yelling at the trolls diminishes you and encourages the trolls. It's a lose-lose proposition. -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
Reply by ●September 25, 20072007-09-25
Vladimir Vassilevsky wrote: ...> КГ/AM (KG/AM): > "Concept is shit/author is dull fuck". > > Слишкам многа букаф: нечо нипонил. > Too many characters; understood nothing. > > Выпей иаду. > Drink a poison. > > What should be the proper English translation?It's perfectly clear as is. Colloquially, the last would become "Drop dead." How come the characters show now? I haven't changed any settings. Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Reply by ●September 25, 20072007-09-25
Jerry Avins wrote:>> =D0=92=D1=8B=D0=BF=D0=B5=D0=B9 =D0=B8=D0=B0=D0=B4=D1=83. >> Drink a poison. >> >> What should be the proper English translation? >=20 > It's perfectly clear as is. Colloquially, the last would become "Drop=20 > dead."Close, but not quite. Depending on the context, poison could also mean=20 the alcohol.> How come the characters show now? I haven't changed any settings.Neither did I. That may have to do with the character encoding set in=20 the message header, the way it is processed by server and understood by=20 the news reader. Mine is set to UTF-8. VLV
Reply by ●September 25, 20072007-09-25
Vladimir Vassilevsky wrote:> > > Jerry Avins wrote: > >>> Выпей иаду. >>> Drink a poison. >>> >>> What should be the proper English translation? >> >> It's perfectly clear as is. Colloquially, the last would become "Drop >> dead." > > Close, but not quite. Depending on the context, poison could also mean > the alcohol. > >> How come the characters show now? I haven't changed any settings. > > Neither did I. That may have to do with the character encoding set in > the message header, the way it is processed by server and understood by > the news reader. Mine is set to UTF-8.Mine too. "It's a puzzlement!" Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
