16 QAM vs 16 PSK and 4QAM vs QPSK

Started by February 15, 2007
Hello All,

Just curious why in many standards, to achieve different rates,
usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or
16-QPSK instead of 16 QAM?

Thanks.

Regards,
Terry

suterr@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM?
If you want to use a constant power output transmitter (e.g. a simple class-C design), QAM doesn't work. You are stuck with modulations that have fixed amplitude symbols - e.g. PSK. However, if you have a linear transmitter, QAM can give better performance, even for small constellations. Regards, Steve
On Feb 15, 9:27 am, sut...@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM?
4-QAM is the same as QPSK. Square m-QAM is normally easier to generate and demodulate (at baseband) than m-PSK, as m-QAM can simply be formed from two orthogonal ASK signals, whereas in m-PSK, one needs angle detection, etc. However, Steve also makes a good argument in favour of PSK, as it has a better PAPR (peak-to-average power ratio). As with any aspect of engineering, any choice is based on compromises. -- Oli -- Oli
Oli Charlesworth wrote:
> On Feb 15, 9:27 am, sut...@gmail.com wrote: >> Hello All, >> >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM? > > 4-QAM is the same as QPSK.
Not necessarily. Haven't you seen 4-QAM where two of the points have a different amplitude from the other two?
> > Square m-QAM is normally easier to generate and demodulate (at > baseband) than m-PSK, as m-QAM can simply be formed from two > orthogonal ASK signals, whereas in m-PSK, one needs angle detection, > etc. > > However, Steve also makes a good argument in favour of PSK, as it has > a better PAPR (peak-to-average power ratio). > > As with any aspect of engineering, any choice is based on compromises.
Regards, Steve
On Feb 15, 11:38 am, Steve Underwood <ste...@dis.org> wrote:
> Oli Charlesworth wrote: > > On Feb 15, 9:27 am, sut...@gmail.com wrote: > >> Hello All, > > >> Just curious why in many standards, to achieve different rates, > >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK
or
> >> 16-QPSK instead of 16 QAM? > > > 4-QAM is the same as QPSK. > > Not necessarily. Haven't you seen 4-QAM where two of the points have a > different amplitude from the other two?
Yes, perhaps I should have said "symmetrical". Although I think that without any other qualifiers, most people would implictly interpret "4-QAM" as the symmetrical type. -- Oli

Steve Underwood wrote:

> suterr@gmail.com wrote: > >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM?
> If you want to use a constant power output transmitter (e.g. a simple > class-C design), QAM doesn't work. You are stuck with modulations that > have fixed amplitude symbols - e.g. PSK.
Due to the bandwidth limitation, PSK does have the nonconstant envelope also.
> However, if you have a linear > transmitter, QAM can give better performance, even for small > constellations.
Actually, the QAM has the advantage over PSK starting from the 8-point constellation. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
suterr@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM? > > Thanks. > > Regards, > Terry >
4-QAM, if it's regular, is the same as QPSK. Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 of the maximum amplitude, and clever but twisted people can probably do better with non-regular constellations. 16-PSK, on the other hand, has a constellation who's closest spacing is about 1/2.5 of the maximum amplitude, for a 7dB penalty in symbol error rate. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Posting from Google? See http://cfaj.freeshell.org/google/ "Applied Control Theory for Embedded Systems" came out in April. See details at http://www.wescottdesign.com/actfes/actfes.html

Tim Wescott wrote:

> Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 > of the maximum amplitude,
What is more interesting is the average spacing vs RMS amplitude.
> and clever but twisted people can probably do > better with non-regular constellations.
Not much better, though. For 16 points, one can gain somewhat 0.5dB advantage over the square constellation. The more points are in the constellation, the less is the difference. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
On Thu, 15 Feb 2007 08:07:30 -0800, Tim Wescott <tim@seemywebsite.com>
wrote:

>suterr@gmail.com wrote: >> Hello All, >> >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM? >> >> Thanks. >> >> Regards, >> Terry >> >4-QAM, if it's regular, is the same as QPSK. > >Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 >of the maximum amplitude, and clever but twisted people can probably do >better with non-regular constellations. 16-PSK, on the other hand, has >a constellation who's closest spacing is about 1/2.5 of the maximum >amplitude, for a 7dB penalty in symbol error rate.
I think that's the short answer, as was alluded to previously, 16-QAM has a higher capacity than 16-PSK. Usually the standards will go toward whatever will provide the best cost/performance tradeoff under the expected impairments, and there's seldom reason to pick 16-PSK over 16-QAM. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org
On Thu, 15 Feb 2007 08:07:30 -0800, Tim Wescott <tim@seemywebsite.com>
wrote:

>suterr@gmail.com wrote: >> Hello All, >> >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM? >> >> Thanks. >> >> Regards, >> Terry >> >4-QAM, if it's regular, is the same as QPSK. > >Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 >of the maximum amplitude, and clever but twisted people can probably do >better with non-regular constellations. 16-PSK, on the other hand, has >a constellation who's closest spacing is about 1/2.5 of the maximum >amplitude, for a 7dB penalty in symbol error rate.
I think that's the short answer, as was alluded to previously, 16-QAM has a higher capacity than 16-PSK. Usually the standards will go toward whatever will provide the best cost/performance tradeoff under the expected impairments, and there's seldom reason to pick 16-PSK over 16-QAM. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org

Tim Wescott wrote:

> Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 > of the maximum amplitude,
What is more interesting is the average spacing vs RMS amplitude.
> and clever but twisted people can probably do > better with non-regular constellations.
Not much better, though. For 16 points, one can gain somewhat 0.5dB advantage over the square constellation. The more points are in the constellation, the less is the difference. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
suterr@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM? > > Thanks. > > Regards, > Terry >
4-QAM, if it's regular, is the same as QPSK. Regular 16-QAM has a constellation who's closest spacing is about 1/5.7 of the maximum amplitude, and clever but twisted people can probably do better with non-regular constellations. 16-PSK, on the other hand, has a constellation who's closest spacing is about 1/2.5 of the maximum amplitude, for a 7dB penalty in symbol error rate. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Posting from Google? See http://cfaj.freeshell.org/google/ "Applied Control Theory for Embedded Systems" came out in April. See details at http://www.wescottdesign.com/actfes/actfes.html

Steve Underwood wrote:

> suterr@gmail.com wrote: > >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM?
> If you want to use a constant power output transmitter (e.g. a simple > class-C design), QAM doesn't work. You are stuck with modulations that > have fixed amplitude symbols - e.g. PSK.
Due to the bandwidth limitation, PSK does have the nonconstant envelope also.
> However, if you have a linear > transmitter, QAM can give better performance, even for small > constellations.
Actually, the QAM has the advantage over PSK starting from the 8-point constellation. Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
On Feb 15, 11:38 am, Steve Underwood <ste...@dis.org> wrote:
> Oli Charlesworth wrote: > > On Feb 15, 9:27 am, sut...@gmail.com wrote: > >> Hello All, > > >> Just curious why in many standards, to achieve different rates, > >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK
or
> >> 16-QPSK instead of 16 QAM? > > > 4-QAM is the same as QPSK. > > Not necessarily. Haven't you seen 4-QAM where two of the points have a > different amplitude from the other two?
Yes, perhaps I should have said "symmetrical". Although I think that without any other qualifiers, most people would implictly interpret "4-QAM" as the symmetrical type. -- Oli
Oli Charlesworth wrote:
> On Feb 15, 9:27 am, sut...@gmail.com wrote: >> Hello All, >> >> Just curious why in many standards, to achieve different rates, >> usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or >> 16-QPSK instead of 16 QAM? > > 4-QAM is the same as QPSK.
Not necessarily. Haven't you seen 4-QAM where two of the points have a different amplitude from the other two?
> > Square m-QAM is normally easier to generate and demodulate (at > baseband) than m-PSK, as m-QAM can simply be formed from two > orthogonal ASK signals, whereas in m-PSK, one needs angle detection, > etc. > > However, Steve also makes a good argument in favour of PSK, as it has > a better PAPR (peak-to-average power ratio). > > As with any aspect of engineering, any choice is based on compromises.
Regards, Steve
On Feb 15, 9:27 am, sut...@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM?
4-QAM is the same as QPSK. Square m-QAM is normally easier to generate and demodulate (at baseband) than m-PSK, as m-QAM can simply be formed from two orthogonal ASK signals, whereas in m-PSK, one needs angle detection, etc. However, Steve also makes a good argument in favour of PSK, as it has a better PAPR (peak-to-average power ratio). As with any aspect of engineering, any choice is based on compromises. -- Oli -- Oli
suterr@gmail.com wrote:
> Hello All, > > Just curious why in many standards, to achieve different rates, > usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or > 16-QPSK instead of 16 QAM?
If you want to use a constant power output transmitter (e.g. a simple class-C design), QAM doesn't work. You are stuck with modulations that have fixed amplitude symbols - e.g. PSK. However, if you have a linear transmitter, QAM can give better performance, even for small constellations. Regards, Steve
Hello All,

Just curious why in many standards, to achieve different rates,
usually use QPSK and 16-QAM. Why can't we use 4-QAM instead of QPSK or
16-QPSK instead of 16 QAM?

Thanks.

Regards,
Terry