Hi, I understand that zero prefix reduces the transmit power in OFDM based systems. Then, why is cyclic prefix(CP) used widely and is more popular ? Why has WiMAX opted for Cyclic Prefix ? Any ideas . Thx in advans, Karthik Balaguru
Zero Prefix and the CP
Started by ●January 18, 2008
Reply by ●January 18, 20082008-01-18
On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru <karthikbalaguru79@gmail.com> wrote:>Hi, > >I understand that zero prefix reduces the transmit power in OFDM based >systems. Then, why is cyclic prefix(CP) used widely and is more >popular ? >Why has WiMAX opted for Cyclic Prefix ? >Any ideas . > >Thx in advans, >Karthik BalaguruCould you clarify what you mean by zero prefix? Do you mean not transmitting any energy during the guard interval? If that's what you mean I don't know of any OFDM systems that work that way. If that's not what you mean, please clarify. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org
Reply by ●January 18, 20082008-01-18
On Jan 18, 10:31�pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote:> On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru > > <karthikbalagur...@gmail.com> wrote: > >Hi, > > >I understand that zero prefix reduces the transmit power in OFDM based > >systems. Then, why is cyclic prefix(CP) used widely and is more > >popular ? > >Why has WiMAX opted for Cyclic Prefix ? > >Any ideas . > > >Thx in advans, > >Karthik Balaguru > > Could you clarify what you mean by zero prefix? � Do you mean not > transmitting any energy during the guard interval? > > If that's what you mean I don't know of any OFDM systems that work > that way. � If that's not what you mean, please clarify. >I find that, Zero prefix constitutes a null guard band. Futher, here, the transmitter will be sending a prefix of null data. That is - send nothing in the guard interval and at the receiver end, copy to the front of OFDM symbol, the received tail. But, I find its use in Ultrawideband (UWB) multiband OFDM. Also refer, http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=4577&objectType=file Any ideas - why Zero Prefix is not preferred in WiMAX ? Karthik Balaguru
Reply by ●January 18, 20082008-01-18
On Jan 18, 6:18 pm, karthikbalaguru <karthikbalagur...@gmail.com> wrote:> On Jan 18, 10:31 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: > > > > > On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru > > > <karthikbalagur...@gmail.com> wrote: > > >Hi, > > > >I understand that zero prefix reduces the transmit power in OFDM based > > >systems. Then, why is cyclic prefix(CP) used widely and is more > > >popular ? > > >Why has WiMAX opted for Cyclic Prefix ? > > >Any ideas . > > > >Thx in advans, > > >Karthik Balaguru > > > Could you clarify what you mean by zero prefix? Do you mean not > > transmitting any energy during the guard interval? > > > If that's what you mean I don't know of any OFDM systems that work > > that way. If that's not what you mean, please clarify. > > I find that, Zero prefix constitutes a null guard band. > Futher, here, the transmitter will be sending a prefix of null data. > That is - send nothing in the guard interval and at the > receiver end, copy to the front of OFDM symbol, the received tail. > > But, I find its use in Ultrawideband (UWB) multiband OFDM. > Also refer,http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objec... > > Any ideas - why Zero Prefix is not preferred in WiMAX ?Because using a cyclic-prefix makes each OFDM easy to equalise; the whole circular-convolution thing. -- Oli
Reply by ●January 18, 20082008-01-18
On Jan 18, 11:53�pm, Oli Charlesworth <ca...@olifilth.co.uk> wrote:> On Jan 18, 6:18 pm, karthikbalaguru <karthikbalagur...@gmail.com> > wrote: > > > > > > > On Jan 18, 10:31 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: > > > > On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru > > > > <karthikbalagur...@gmail.com> wrote: > > > >Hi, > > > > >I understand that zero prefix reduces the transmit power in OFDM based > > > >systems. Then, why is cyclic prefix(CP) used widely and is more > > > >popular ? > > > >Why has WiMAX opted for Cyclic Prefix ? > > > >Any ideas . > > > > >Thx in advans, > > > >Karthik Balaguru > > > > Could you clarify what you mean by zero prefix? � Do you mean not > > > transmitting any energy during the guard interval? > > > > If that's what you mean I don't know of any OFDM systems that work > > > that way. � If that's not what you mean, please clarify. > > > I find that, Zero prefix constitutes a null guard band. > > Futher, here, the transmitter will be sending a prefix of null data. > > That is - send nothing in the guard interval and at the > > receiver end, copy to the front of OFDM symbol, the received tail. > > > But, I find its use in Ultrawideband (UWB) multiband OFDM. > > Also refer,http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objec... > > > Any ideas - why Zero Prefix is not preferred in WiMAX ? > > Because using a cyclic-prefix makes each OFDM easy to equalise; the > whole circular-convolution thing. >Thx for that info !! But, i still have some queries based on this . It looks like, zero prefix reduces the transmit power in OFDM based systems. Transmit power also plays an important role. So, why is cyclic prefix(CP) used instead of Zero Prefix ? Any other reasons ? Karthik Balaguru
Reply by ●January 18, 20082008-01-18
karthikbalaguru wrote:> On Jan 18, 11:53 pm, Oli Charlesworth <ca...@olifilth.co.uk> wrote: >> On Jan 18, 6:18 pm, karthikbalaguru <karthikbalagur...@gmail.com> >> wrote: >> >> >> >> >> >>> On Jan 18, 10:31 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: >>>> On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru >>>> <karthikbalagur...@gmail.com> wrote: >>>>> Hi, >>>>> I understand that zero prefix reduces the transmit power in OFDM based >>>>> systems. Then, why is cyclic prefix(CP) used widely and is more >>>>> popular ? >>>>> Why has WiMAX opted for Cyclic Prefix ? >>>>> Any ideas . >>>>> Thx in advans, >>>>> Karthik Balaguru >>>> Could you clarify what you mean by zero prefix? Do you mean not >>>> transmitting any energy during the guard interval? >>>> If that's what you mean I don't know of any OFDM systems that work >>>> that way. If that's not what you mean, please clarify. >>> I find that, Zero prefix constitutes a null guard band. >>> Futher, here, the transmitter will be sending a prefix of null data. >>> That is - send nothing in the guard interval and at the >>> receiver end, copy to the front of OFDM symbol, the received tail. >>> But, I find its use in Ultrawideband (UWB) multiband OFDM. >>> Also refer,http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objec... >>> Any ideas - why Zero Prefix is not preferred in WiMAX ? >> Because using a cyclic-prefix makes each OFDM easy to equalise; the >> whole circular-convolution thing. >> > > Thx for that info !! But, i still have some queries based on this . > It looks like, zero prefix reduces the transmit power in OFDM based > systems. > Transmit power also plays an important role. > So, why is cyclic prefix(CP) used instead of Zero Prefix ? > Any other reasons ?In the end, it's bit error rate that matters. Turning the transmitter off saves power, but it doesn't help send good bits through. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●January 18, 20082008-01-18
On Fri, 18 Jan 2008 10:18:04 -0800 (PST), karthikbalaguru <karthikbalaguru79@gmail.com> wrote:>On Jan 18, 10:31�pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: >> On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru >> >> <karthikbalagur...@gmail.com> wrote: >> >Hi, >> >> >I understand that zero prefix reduces the transmit power in OFDM based >> >systems. Then, why is cyclic prefix(CP) used widely and is more >> >popular ? >> >Why has WiMAX opted for Cyclic Prefix ? >> >Any ideas . >> >> >Thx in advans, >> >Karthik Balaguru >> >> Could you clarify what you mean by zero prefix? � Do you mean not >> transmitting any energy during the guard interval? >> >> If that's what you mean I don't know of any OFDM systems that work >> that way. � If that's not what you mean, please clarify. >> > >I find that, Zero prefix constitutes a null guard band. >Futher, here, the transmitter will be sending a prefix of null data. >That is - send nothing in the guard interval and at the >receiver end, copy to the front of OFDM symbol, the received tail.>But, I find its use in Ultrawideband (UWB) multiband OFDM. >Also refer, http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=4577&objectType=fileI had to do some reading to catch up, but I think I get it now. Yes, it will require less total transmit power, and since a goal for UWB and 802.15 devices in general is low power consumption, I suppose that helps a little bit. According to one reference I found it may also reduce ripple in the Tx spectrum a bit, but the value of that may be questionable in many cases. UWB has some pretty tight psd limits, so perhaps that was part of the motivation.>Any ideas - why Zero Prefix is not preferred in WiMAX ? > >Karthik BalaguruThere are some significant downsides to Zero Prefix for a lot of applications. PAPR is problematic for OFDM in general, and turning the Power Amplifier off for a small amount of time in a high-power application is generally pretty difficult. The power levels in UWB are quite low since the range is miniscule compared to WiMAX, and the PA technology for UWB is evidently tolerant of the on-off cycles. Getting a WiMAX PA to do that, or any PA that's transmitting very much power, is much harder to do. Controlling the PA on/off cycle is important in keeping it from splattering energy into adjacent bands, and turning it on and off every OFDM symbol would create a lot more problems than it solves. That's my guess, anyway. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org
Reply by ●January 18, 20082008-01-18
"karthikbalaguru" <karthikbalaguru79@gmail.com> wrote in message news:5397de59-cac2-4254-9f7e-cb55ff92305f@s8g2000prg.googlegroups.com...> Hi, > > I understand that zero prefix reduces the transmit power in OFDM based > systems. Then, why is cyclic prefix(CP) used widely and is more > popular ?Think what happens if there is a dispersion in the channel. Vladimir Vassilevsky DSP and Mixed Signal Consultant www.abvolt.com
Reply by ●January 18, 20082008-01-18
The cyclic prefix is essential. Otherwise the dispersion in the channel will result in the loss of the orthogonality. VLV "Eric Jacobsen" <eric.jacobsen@ieee.org> wrote in message news:o9d2p31hhho7oas00n9svmembcp20o67bn@4ax.com...> On Fri, 18 Jan 2008 10:18:04 -0800 (PST), karthikbalaguru > <karthikbalaguru79@gmail.com> wrote: > > >On Jan 18, 10:31 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: > >> On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru > >> > >> <karthikbalagur...@gmail.com> wrote: > >> >Hi, > >> > >> >I understand that zero prefix reduces the transmit power in OFDM based > >> >systems. Then, why is cyclic prefix(CP) used widely and is more > >> >popular ? > >> >Why has WiMAX opted for Cyclic Prefix ? > >> >Any ideas . > >> > >> >Thx in advans, > >> >Karthik Balaguru > >> > >> Could you clarify what you mean by zero prefix? Do you mean not > >> transmitting any energy during the guard interval? > >> > >> If that's what you mean I don't know of any OFDM systems that work > >> that way. If that's not what you mean, please clarify. > >> > > > >I find that, Zero prefix constitutes a null guard band. > >Futher, here, the transmitter will be sending a prefix of null data. > >That is - send nothing in the guard interval and at the > >receiver end, copy to the front of OFDM symbol, the received tail. > > >But, I find its use in Ultrawideband (UWB) multiband OFDM. > >Also refer,http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=4577&objectType=file> > I had to do some reading to catch up, but I think I get it now. > > Yes, it will require less total transmit power, and since a goal for > UWB and 802.15 devices in general is low power consumption, I suppose > that helps a little bit. According to one reference I found it may > also reduce ripple in the Tx spectrum a bit, but the value of that may > be questionable in many cases. UWB has some pretty tight psd limits, > so perhaps that was part of the motivation. > > >Any ideas - why Zero Prefix is not preferred in WiMAX ? > > > >Karthik Balaguru > > There are some significant downsides to Zero Prefix for a lot of > applications. PAPR is problematic for OFDM in general, and turning > the Power Amplifier off for a small amount of time in a high-power > application is generally pretty difficult. The power levels in UWB > are quite low since the range is miniscule compared to WiMAX, and the > PA technology for UWB is evidently tolerant of the on-off cycles. > Getting a WiMAX PA to do that, or any PA that's transmitting very much > power, is much harder to do. Controlling the PA on/off cycle is > important in keeping it from splattering energy into adjacent bands, > and turning it on and off every OFDM symbol would create a lot more > problems than it solves. > > That's my guess, anyway. > > Eric Jacobsen > Minister of Algorithms > Abineau Communications > http://www.ericjacobsen.org
Reply by ●January 19, 20082008-01-19
Vlad, the zero prefix doesn't eliminate the prefix, it replaces it with dead time. i.e., transmit no energy during the CP time. In the receiver the dead time after the symbol is sampled, so that the tail due to the channel memory can be added back to the beginning of the FFT vector so that the channel convolution is still circular. It does mean that you're not transmitting any energy during the CP, it also means that the PA is turning OFF for the CP duration every symbol. Depending on how the PA is made that could be good for energy consumption and spectral flattening or it could royally screw up the PA behavior. On Sat, 19 Jan 2008 02:52:40 GMT, "Vladimir Vassilevsky" <antispam_bogus@hotmail.com> wrote:> >The cyclic prefix is essential. Otherwise the dispersion in the channel will >result in the loss of the orthogonality. > >VLV > > > >"Eric Jacobsen" <eric.jacobsen@ieee.org> wrote in message >news:o9d2p31hhho7oas00n9svmembcp20o67bn@4ax.com... >> On Fri, 18 Jan 2008 10:18:04 -0800 (PST), karthikbalaguru >> <karthikbalaguru79@gmail.com> wrote: >> >> >On Jan 18, 10:31 pm, Eric Jacobsen <eric.jacob...@ieee.org> wrote: >> >> On Fri, 18 Jan 2008 09:16:45 -0800 (PST), karthikbalaguru >> >> >> >> <karthikbalagur...@gmail.com> wrote: >> >> >Hi, >> >> >> >> >I understand that zero prefix reduces the transmit power in OFDM based >> >> >systems. Then, why is cyclic prefix(CP) used widely and is more >> >> >popular ? >> >> >Why has WiMAX opted for Cyclic Prefix ? >> >> >Any ideas . >> >> >> >> >Thx in advans, >> >> >Karthik Balaguru >> >> >> >> Could you clarify what you mean by zero prefix? Do you mean not >> >> transmitting any energy during the guard interval? >> >> >> >> If that's what you mean I don't know of any OFDM systems that work >> >> that way. If that's not what you mean, please clarify. >> >> >> > >> >I find that, Zero prefix constitutes a null guard band. >> >Futher, here, the transmitter will be sending a prefix of null data. >> >That is - send nothing in the guard interval and at the >> >receiver end, copy to the front of OFDM symbol, the received tail. >> >> >But, I find its use in Ultrawideband (UWB) multiband OFDM. >> >Also refer, >http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=4577&objectType=file >> >> I had to do some reading to catch up, but I think I get it now. >> >> Yes, it will require less total transmit power, and since a goal for >> UWB and 802.15 devices in general is low power consumption, I suppose >> that helps a little bit. According to one reference I found it may >> also reduce ripple in the Tx spectrum a bit, but the value of that may >> be questionable in many cases. UWB has some pretty tight psd limits, >> so perhaps that was part of the motivation. >> >> >Any ideas - why Zero Prefix is not preferred in WiMAX ? >> > >> >Karthik Balaguru >> >> There are some significant downsides to Zero Prefix for a lot of >> applications. PAPR is problematic for OFDM in general, and turning >> the Power Amplifier off for a small amount of time in a high-power >> application is generally pretty difficult. The power levels in UWB >> are quite low since the range is miniscule compared to WiMAX, and the >> PA technology for UWB is evidently tolerant of the on-off cycles. >> Getting a WiMAX PA to do that, or any PA that's transmitting very much >> power, is much harder to do. Controlling the PA on/off cycle is >> important in keeping it from splattering energy into adjacent bands, >> and turning it on and off every OFDM symbol would create a lot more >> problems than it solves. >> >> That's my guess, anyway. >> >> Eric Jacobsen >> Minister of Algorithms >> Abineau Communications >> http://www.ericjacobsen.org >Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org
