"Carlos Moreno" <moreno_at_mochima_dot_com@mailinator.com> wrote in message 
news:rS6xf.68040$2O3.1269092@wagner.videotron.net...
> ravirevolt wrote:
>> Hi I'm new to DSP and i'm doing a project on simulation of Digital
>> Beamforming. I want to know how to get started with this. These are the
>> objectives of the project :
>> 1: To receive the RF signal, process and digitize the signal, and 
>> generate
>> multiple receive beams.
>> 2: To incorporate variable number of array elements and beams
>> 3: To steer the generated beams in specific directions
>> 4: To restrict sidelobe levels as sepecified
>> please help me on this. I'm a computer science student. what are the 
>> steps
>> to be followed in doing this. How can i get started. help me.
>
> I don't understand -- does step 1 mean that you have to actually work
> with a real signal?  Or are *all* of the above steps to be simulated
> by software?  Also, digitize an *RF* signal?  I really doubt it (the
> frequency is just too high)
>
> If we're talking 100% simulation, you have to start by setting up a
> function of space-time corresponding to a moving wave -- in the one-
> dimensional case, we're talking about a sin (w*t - l*x) kind of
> function.  In 2- or 3-dimensional space, you have to generalize it.
>
> Then, simply "sample" the signal at the right points (by taking the
> value of the function at the given time and at the given points in
> space) -- after that, you're all set;  there's a lot of design of
> the beamforming parameters to be done (to accomplish step 3 and 4,
> at least).
>
> HTH,
>
> Carlos
What Carlos says.  Basically, if your sidelobe spec allows it you will 
probably get away with assuming point isotropic radiators, ignoring 
polarisation and mutual coupling and looking only at regularly spaced linear 
arrays of elements - this ought to significantly simplify your problem and 
you can always complicate it later if you want. You also need signal center 
frequency and bandwidth so you can do an adequate job of representing the 
sampled baseband component before messing it up in an imperfect digital beam 
former.  You might find that building an analytical model will give you a 
good initial insight , it should also give you a nice tool for deciding what 
your minimum element spacing should be and for plotting 'ideal' patterns to 
compare your digital beamformer patterns against.

Best of Luck - Mike
P.S. If you make all of your combinatorial logic delay based (ensure that 
any interpolation is linear phase shift with frequency over the full 
bandwidth occupied by any of your signals ) you may end up with a 
computationally expensive model but it's so much simpler to work out what's 
going on.

ravirevolt wrote:
> Hi I'm new to DSP and i'm doing a project on simulation of Digital
> Beamforming. I want to know how to get started with this. These are the
> objectives of the project :
> 1: To receive the RF signal, process and digitize the signal, and generate
> multiple receive beams.
> 2: To incorporate variable number of array elements and beams
> 3: To steer the generated beams in specific directions
> 4: To restrict sidelobe levels as sepecified
> please help me on this. I'm a computer science student. what are the steps
> to be followed in doing this. How can i get started. help me.

I don't understand -- does step 1 mean that you have to actually work
with a real signal?  Or are *all* of the above steps to be simulated
by software?  Also, digitize an *RF* signal?  I really doubt it (the
frequency is just too high)

If we're talking 100% simulation, you have to start by setting up a
function of space-time corresponding to a moving wave -- in the one-
dimensional case, we're talking about a sin (w*t - l*x) kind of
function.  In 2- or 3-dimensional space, you have to generalize it.

Then, simply "sample" the signal at the right points (by taking the
value of the function at the given time and at the given points in
space) -- after that, you're all set;  there's a lot of design of
the beamforming parameters to be done (to accomplish step 3 and 4,
at least).

HTH,

Carlos
--

Hi I'm new to DSP and i'm doing a project on simulation of Digital
Beamforming. I want to know how to get started with this. These are the
objectives of the project :
1: To receive the RF signal, process and digitize the signal, and generate
multiple receive beams.
2: To incorporate variable number of array elements and beams
3: To steer the generated beams in specific directions
4: To restrict sidelobe levels as sepecified
please help me on this. I'm a computer science student. what are the steps
to be followed in doing this. How can i get started. help me.