matlab antenna array operations
the example:
- radiation patern
- gain of traveling-wave antenna of length L
- simplified Yagi-Uda array design
- etc
wish u need it :D
matlab is too slow, especially when the matrix is very large.
maybe fortran is much better.
I wrote a code about Antenna array using MatLab... I've to expand it for what concerns the Optimization and I'm new in this field.. Can u suggest some optimization function?
In other words, I've a function computing the Array Pattern...given a mask, I can define an error function (obvious).. can I find an optimization function nulling the error function changing the inputs to the Array Pattern function?
Thanks in advance
Lupin
Matlab is optimized for matrix operations. I would doubt that anyone starting on their own could write an improved method.
I'm not too familier with fortran.
yea...may be Fortran is better for matrix operation, but I think Matlab is more simple to write program .
i guess supernec is better in design and simulation settings it's very fast and good in design especially in yagi antennas
Lord, Can u give me information about supernec.
It's a free software or whatever?, i just know it because I guess supernec is not too universal than matlab.
superNEC is a very expensive program that only analyzes antennas. They put a shell around Excel. [edited part. I misremembered from another program. It is matlab as eirp says.]
As far as matlab, their routines to do matrix math are very optimized and take advantage of zeros in the array.
Hi all..
Supernec (current ver 2.55) is a very nice MoM-UTD 3rd party application for Matlab (Win and Linux).
It's very accurate for wire antennas but quite unfriendly :)
Advantage is using theory of diffraction together with MoM for large objects..
http://www.poynting.co.za/software/snec_detail.shtml
I think matlab is simple than fortran!!! But it is slower! Can anyone give me some advice???
Matlab is an interpreted language. This means that the source code is not compiled but is interpreted on the fly. Because Matlab is an interpreted language, it must interpret every line in a loop each time it goes through the loop. This makes Matlab painfully slow at doing loops.
So I guess Fortran/C would be your best bet for creating faster matrix inversions.You should exploit the symmetry structures in you Matrix.For example inverting a toeplitz matrix requires far less time than a general matrix.You should look at many available algorithms for doing fast matrix inversion.May be you could try assembly code if you are brave enough !!!
Matlab is very useful for solve matrix and plot image than fortran.Althought it is slow,when we solve small and not complex problem matlab is like a "draft"