Hello all

I have started work on side scan sonar modeling and simulations. Currently i am just working on simple linear array beam forming. As a startup, i have simulated linear array sensor's data (time domain signal) using MATLAB phased array toolbox and verified different types of beam former using this array data. The basic theme behind it is to incorporate corresponding delays/phase shifts in sensors data depending upon direction of arrival of incoming signal/echo.

Now i need to simulate shadow of a bottom laying object placed at at specific bearing & range w.r.t array at array sensors level (in time domain) so that when i apply this data at beam former input,  i can get corresponding shadow of that object in beam former output display.

Can any one kindly help how to simulate shadow's effect at array's sensor level?

Note: My English is not that good so i apologize in advance if i could not describe my problem clearly.

Thanks 

Nauman

#Beamforming

Can you use a ray tracing method?  Go from the reflection points back towards the sensor and determine if it is blocked.  If you do enough rays you will get an idea of how much of the object prevents reflected signal getting back (that's what I think you mean by "shadow").  It's the same idea used in 3D image graphics, you go from the object to the eye to determine if that object is actually visible from the eye's perspective.

Dr. mike

Dear Dr. Mike

Thanks for help. I have used MATLAB based acoustic toolbox some time ago which uses bellhop and other ray tracing models for transmission loss calculations. Among other problems, one immediate problem i think i face is how to convert ray tracing model output into array sensor's time domain output?

Can you kindly explain/point out some literature useful in this regard?

Thanks

Nauman

I just did a web search using "sonar ray trace" and got a lot of nice papers:

http://link.springer.com/chapter/10.1007/0-387-234...

http://www.researchgate.net/profile/Jens_Hovem/pub...

This one looks really old:

http://faculty.nps.edu/brutzman/vrtp/rra/rra.html

so I would try to follow that forward using citation indexing.

One of the phrases I saw was "this is an unsolved problem", so it's not so easy to do what you are attempting.  But I suspect you can at least get a reasonable simulation.  I have never attempted anything like this, so I can't really help.  It just seems like a fun problem to work on!

Dr. mike

Thanks Dr. Mike for lterature