Hi dsp/wireless gurus,
please answer my previous questions, i am really stuck at those
points..will be thankfull to you dsp gurus for doing helping me out.
meanwhile i got one more fresh doubt after reading pasupathy paper of
[ MINIMUM SHIFT KEYING:A SPECTRALL\r EFFICIENT MODULATION ] by
subbarayan pasupathy 1979.
in that paper he explains about few diagrams and it means the
following
a +ve half sinusoid represents 1 and -ve half sinusoid respresents -1
[bipolar format is considered here]
he multiplies a carrier which is COSINE of one full cycle so
essentially
msk +ve pulse = (one +ve half sinusoid is having width of T time
units) * (one COSINE full cyle having one full cycle within T time
units)
diagram of +ve half sinusoid occupying T time units
_
/ \
/ \
/ \
/ \
/ \
/ \
/-------------------\--
<- -----T--------->
Assuming 8 Samples per Symbol
problem situation: - i have a carrier (assume COSINE) of 2 GHz and 1
cycle corresponds to 1/(2e9) time units. i may not have data at that
rate ..i mean data bit having 1/(2e9) time units ..say i have data
rate at 5MHz so 1 data bit may stretch 1/(5e6) time units. In this
case each DATA BIT
undergoes multiplication with 400 COSINE full cycles
|---------------------------------------------------------------------------------------1
data bit-------------------------------------------|
MSK = | cosine full cycle1/cosine full cycle2/cosine
full cycle3..................................cosine full cycle400 |
(2e9)/(5e6) = 400.
questions
1. is my understanding correct.
2. is the abrupt phase change still contained (i mean is the purpose
of MSK still maintained..i meant will there wont be sudden abrupt
phase changes for the final MSK symbol.
Thanks
A.Ahmed
MSK and its generation
Started by ●June 26, 2011
Reply by ●June 26, 20112011-06-26
On Sun, 26 Jun 2011 06:39:58 -0700, aizza ahmed wrote:> Hi dsp/wireless gurus, > please answer my previous questions, i am really stuck at those > points..will be thankfull to you dsp gurus for doing helping me out. > meanwhile i got one more fresh doubt after reading pasupathy paper of [ > MINIMUM SHIFT KEYING:A SPECTRALL\r EFFICIENT MODULATION ] by subbarayan > pasupathy 1979. > > in that paper he explains about few diagrams and it means the following > > a +ve half sinusoid represents 1 and -ve half sinusoid respresents -1 > [bipolar format is considered here] > > he multiplies a carrier which is COSINE of one full cycle so > essentially > > msk +ve pulse = (one +ve half sinusoid is having width of T time > units) * (one COSINE full cyle having one full cycle within T time > units) > > diagram of +ve half sinusoid occupying T time units > _ > / \ > / \ > / \ > / \ > / \ > / \ > /-------------------\-- > <- -----T---------> > > Assuming 8 Samples per Symbol > > problem situation: - i have a carrier (assume COSINE) of 2 GHz and 1 > cycle corresponds to 1/(2e9) time units. i may not have data at that > rate ..i mean data bit having 1/(2e9) time units ..say i have data rate > at 5MHz so 1 data bit may stretch 1/(5e6) time units. In this case each > DATA BIT > undergoes multiplication with 400 COSINE full cycles > > > >|---------------------------------------------------------------------------------------1> data bit-------------------------------------------| MSK = > | cosine full cycle1/cosine full cycle2/cosine full > cycle3..................................cosine full cycle400 | > > > (2e9)/(5e6) = 400. > > questions > 1. is my understanding correct. > 2. is the abrupt phase change still contained (i mean is the purpose of > MSK still maintained..i meant will there wont be sudden abrupt phase > changes for the final MSK symbol.I don't have that paper in front of me, but I'm 99.44% sure that he's referring to what you do with the signal at baseband, and that modulating it to a real carrier is left as an exercise to the reader. The beauty of MSK is that you don't have to modulate it as OQPSK -- you can modulate it as FSK, and because the phase shift is just right, it acts like OQPSK. There's more detail here, if you dig for it: http://www.wescottdesign.com/articles/MSK/mskTop.html If you notice that this stuff is mind-bending and hard to understand keep in mind that it's not because you're stupid: it's that way to nearly all of us. -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
Reply by ●June 29, 20112011-06-29
On Jun 26, 11:37�pm, Tim <t...@seemywebsite.please> wrote:> On Sun, 26 Jun 2011 06:39:58 -0700, aizza ahmed wrote: > > Hi dsp/wireless gurus, > > � � � �please answer my previous questions, i am really stuck at those > > points..will be thankfull to you dsp gurus for doing helping me out. > > meanwhile i got one more fresh doubt after reading pasupathy paper of [ > > MINIMUM SHIFT KEYING:A SPECTRALL\r EFFICIENT MODULATION ] by subbarayan > > pasupathy 1979. > > > in that paper he explains about few diagrams and it means the following > > > a +ve half sinusoid represents 1 and -ve half sinusoid respresents -1 > > [bipolar format is considered here] > > > he multiplies a carrier �which is COSINE of one full cycle so > > essentially > > > �msk +ve �pulse = (one +ve half sinusoid is having width of T time > > units) * (one COSINE full cyle having one full cycle within T time > > units) > > > diagram of +ve half sinusoid occupying T time units > > � � � � � _ > > � � � � / � \ > > � � � �/ � � \ > > � � �/ � � � � \ > > � �/ � � � � � � \ > > � / � � � � � � � \ > > �/ � � � � � � � � \ > > /-------------------\-- > > <- -----T---------> > > > Assuming 8 Samples per Symbol > > > �problem situation: - i have a carrier (assume COSINE) of 2 GHz and 1 > > cycle corresponds to 1/(2e9) time units. i may not have data at that > > rate ..i mean data bit having 1/(2e9) time units ..say i have data rate > > at 5MHz so 1 data bit may stretch 1/(5e6) time units. In this case each > > DATA BIT > > undergoes multiplication with 400 COSINE full cycles > > |-------------------------------------------------------------------------- -------------1 > > > data bit-------------------------------------------| MSK = � � � � � � � > > � | cosine full cycle1/cosine full cycle2/cosine full > > cycle3..................................cosine full cycle400 � | > > > (2e9)/(5e6) = 400. > > > questions > > 1. is my understanding correct. > > 2. is the abrupt phase change still contained (i mean is the purpose of > > MSK still maintained..i meant will there wont be sudden abrupt phase > > changes for the final MSK symbol. > > I don't have that paper in front of me, but I'm 99.44% sure that he's > referring to what you do with the signal at baseband, and that modulating > it to a real carrier is left as an exercise to the reader. > > The beauty of MSK is that you don't have to modulate it as OQPSK -- you > can modulate it as FSK, and because the phase shift is just right, it > acts like OQPSK. > > There's more detail here, if you dig for it:http://www.wescottdesign.com/articles/MSK/mskTop.html > > If you notice that this stuff is mind-bending and hard to understand keep > in mind that it's not because you're stupid: it's that way to nearly all > of us. > > -- > Tim Wescott > Control system and signal processing consultingwww.wescottdesign.comHi Tim, he has used the symbols something like this ai(t)cos(pi*t/2T) ...here ai is INPHASE amplitude (baseband) next he multiplies with cos(2*pi*fc*t) (RF) so its overall ai(t)cos(pi*t/2T)*cos(2*pi*fc*t) similarly quadrature phase also aq(t)sin(pi*t/2T)*sin(2*pi*fc*t) what he has shown is one cyle of baseband and one cycle of RF..and also one more paper from charan langton on complextoreal.com has similar description http://www.complextoreal.com/chapters/mod1.pdf please let me know any hints/solutions to my questions.. Regards
Reply by ●June 29, 20112011-06-29
On 06/29/2011 10:50 AM, aizza ahmed wrote:> On Jun 26, 11:37 pm, Tim<t...@seemywebsite.please> wrote: >> On Sun, 26 Jun 2011 06:39:58 -0700, aizza ahmed wrote: >>> Hi dsp/wireless gurus, >>> please answer my previous questions, i am really stuck at those >>> points..will be thankfull to you dsp gurus for doing helping me out. >>> meanwhile i got one more fresh doubt after reading pasupathy paper of [ >>> MINIMUM SHIFT KEYING:A SPECTRALL\r EFFICIENT MODULATION ] by subbarayan >>> pasupathy 1979. >> >>> in that paper he explains about few diagrams and it means the following >> >>> a +ve half sinusoid represents 1 and -ve half sinusoid respresents -1 >>> [bipolar format is considered here] >> >>> he multiplies a carrier which is COSINE of one full cycle so >>> essentially >> >>> msk +ve pulse = (one +ve half sinusoid is having width of T time >>> units) * (one COSINE full cyle having one full cycle within T time >>> units) >> >>> diagram of +ve half sinusoid occupying T time units >>> _ >>> / \ >>> / \ >>> / \ >>> / \ >>> / \ >>> / \ >>> /-------------------\-- >>> <- -----T---------> >> >>> Assuming 8 Samples per Symbol >> >>> problem situation: - i have a carrier (assume COSINE) of 2 GHz and 1 >>> cycle corresponds to 1/(2e9) time units. i may not have data at that >>> rate ..i mean data bit having 1/(2e9) time units ..say i have data rate >>> at 5MHz so 1 data bit may stretch 1/(5e6) time units. In this case each >>> DATA BIT >>> undergoes multiplication with 400 COSINE full cycles >> >> |-------------------------------------------------------------------------- -------------1 >> >>> data bit-------------------------------------------| MSK = >>> | cosine full cycle1/cosine full cycle2/cosine full >>> cycle3..................................cosine full cycle400 | >> >>> (2e9)/(5e6) = 400. >> >>> questions >>> 1. is my understanding correct. >>> 2. is the abrupt phase change still contained (i mean is the purpose of >>> MSK still maintained..i meant will there wont be sudden abrupt phase >>> changes for the final MSK symbol. >> >> I don't have that paper in front of me, but I'm 99.44% sure that he's >> referring to what you do with the signal at baseband, and that modulating >> it to a real carrier is left as an exercise to the reader. >> >> The beauty of MSK is that you don't have to modulate it as OQPSK -- you >> can modulate it as FSK, and because the phase shift is just right, it >> acts like OQPSK. >> >> There's more detail here, if you dig for it:http://www.wescottdesign.com/articles/MSK/mskTop.html >> >> If you notice that this stuff is mind-bending and hard to understand keep >> in mind that it's not because you're stupid: it's that way to nearly all >> of us. >> >> -- >> Tim Wescott >> Control system and signal processing consultingwww.wescottdesign.com > > Hi Tim, > he has used the symbols something like this > > ai(t)cos(pi*t/2T) ...here ai is INPHASE amplitude (baseband) > > next he multiplies with cos(2*pi*fc*t) (RF) > > so its overall ai(t)cos(pi*t/2T)*cos(2*pi*fc*t) > > similarly quadrature phase also > > aq(t)sin(pi*t/2T)*sin(2*pi*fc*t) > > what he has shown is one cyle of baseband and one cycle of RF..and > also one more paper from charan langton on complextoreal.com has > similar description > > http://www.complextoreal.com/chapters/mod1.pdf > > please let me know any hints/solutions to my questions..Work out the math. What is cos(pi*t/2T)*cos(2*pi*fc*t) + sin(pi*t/2T)*sin(2*pi*fc*t)? What is cos(pi*t/2T)*cos(2*pi*fc*t) - sin(pi*t/2T)*sin(2*pi*fc*t)? What about the other two combinations? How does this relate to my statement that MSK can be viewed as either OQPSK or FSK? Did you bother to read the pertinent chapter of my thesis? -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" was written for you. See details at http://www.wescottdesign.com/actfes/actfes.html
