DSPRelated.com
Forums

phase of rayleigh envelop

Started by mira...@yahoo.com March 2, 2008
hi
i have written matlab code for rayleigh fading using BPSK and QPSK
i generate a rayleigh distribution by generating two identiaccly independently generated gaussian vectors and than i take a absolute value of it that gives me a rayleigh distribution..than i multiply this value with transmitted signal that gives me a rayleigh fading signal..that means i multiply amplitude of rayleigh envelop with the signal but we are ignoring the phse?
1)so anybody tell me why we ignoring the phase...because it is uniformaly distibuted between 0 to 2 pi..or it is independent or other reason..
2)do we have to consider a phase or not? and why?

i m really confuse at this point..i really appreciate it if anybody can give me appropriate answer
thanking u in anticipation
The question is why are you generate a fading rayleigh distributed?
Probably because you are trying to simulate the propagation channel of a
radio mobile system.
This fading MUST be complex because the propagation channel have effect on
module and phase.
Why the distribution is Rayleight?
If you consider separately the real part and imaginary part of the channel,
and suppose that this effect
is due to a lot of objects around your mobile station, there is the theorem
of the central limit that say
that the process is gaussian distributed.
Now if you analyze the new random variable obtained like sqrt(re^2+im^2) you
can verify that this is
Rayleigh distributed and the formula represent the fading amplitude but
there is always the phase that
you can not neglecting. What I suggest you is verify haw to generate this
fading

On Fri, Feb 29, 2008 at 8:35 PM, wrote:

> hi
> i have written matlab code for rayleigh fading using BPSK and QPSK
> i generate a rayleigh distribution by generating two identiaccly
> independently generated gaussian vectors and than i take a absolute value of
> it that gives me a rayleigh distribution..than i multiply this value with
> transmitted signal that gives me a rayleigh fading signal..that means i
> multiply amplitude of rayleigh envelop with the signal but we are ignoring
> the phse?
> 1)so anybody tell me why we ignoring the phase...because it is uniformaly
> distibuted between 0 to 2 pi..or it is independent or other reason..
> 2)do we have to consider a phase or not? and why?
>
> i m really confuse at this point..i really appreciate it if anybody can
> give me appropriate answer
> thanking u in anticipation
>
>
>
The phase have to be added automatically .
Consider the possibility to generate the fading signal like

f(t) = SUM i=1,N ai exp(-j2pi/lambda*v*t*cos(alphai)

where

ai is a random variable gaussian distributed (mean=0)
v is the speed of the mobile
t is the time of the simulation
aphai is a random variable uniform distributed -pi and pi
N is the number af wave superimposed to produce multipath an then Rayleight
fading (N~10,20)
ai is used to control the power os the fading

if you analyze this signal is amplitude is rayleight distributed and the
phase is correct.
This procedure consider also the Doppler effect present in a channel for
mobile communications

This approach derive from CLARK.

bye

On Mon, Mar 3, 2008 at 6:34 PM, miral patel wrote:

> thanks for your reply ..i really appreciate it..but can u tell me how to
> add a phse of rayleigh faded signal in to transmitted sigal..in my case i
> have to modulate a digital signal transmitted through rayleigh channel and
> demodulate it and plot a graph for bit error rate..i mean there is a some
> procedure to add phase of rayleigh signal to transmitted signal..either can
> u give me books name that can resolve my problem....
>
> thanks agian.... waiting for ur postive reply...
>
> Miral Patel
> Glamorgan University
> UK
> 07809827239
> ----- Original Message ----
> From: arkkimede
> To: m...@yahoo.com
> Cc: m...
> Sent: Monday, March 3, 2008 5:07:36 PM
> Subject: Re: [matlab] phase of rayleigh envelop
>
> The question is why are you generate a fading rayleigh distributed?
> Probably because you are trying to simulate the propagation channel of a
> radio mobile system.
> This fading MUST be complex because the propagation channel have effect on
> module and phase.
> Why the distribution is Rayleight?
> If you consider separately the real part and imaginary part of the
> channel, and suppose that this effect
> is due to a lot of objects around your mobile station, there is the
> theorem of the central limit that say
> that the process is gaussian distributed.
> Now if you analyze the new random variable obtained like sqrt(re^2+im^2)
> you can verify that this is
> Rayleigh distributed and the formula represent the fading amplitude but
> there is always the phase that
> you can not neglecting. What I suggest you is verify haw to generate this
> fading
>
> On Fri, Feb 29, 2008 at 8:35 PM, wrote:
>
> > hi
> > i have written matlab code for rayleigh fading using BPSK and QPSK
> > i generate a rayleigh distribution by generating two identiaccly
> > independently generated gaussian vectors and than i take a absolute value of
> > it that gives me a rayleigh distribution..than i multiply this value with
> > transmitted signal that gives me a rayleigh fading signal..that means i
> > multiply amplitude of rayleigh envelop with the signal but we are ignoring
> > the phse?
> > 1)so anybody tell me why we ignoring the phase...because it is
> > uniformaly distibuted between 0 to 2 pi..or it is independent or other
> > reason..
> > 2)do we have to consider a phase or not? and why?
> >
> > i m really confuse at this point..i really appreciate it if anybody can
> > give me appropriate answer
> > thanking u in anticipation
> >
> >
> > ------------------------------
> Never miss a thing. Make Yahoo your homepage.