correlation coefficient for two dipole antennas

[Jone]

I'm interested in finding the correlation between two (or more) dipole wire antennas as a function of the separation distance. I've trid this using SuperNec (uses MoM), but this software seems to limitied to simpler calculations. Can you recommend some good software based on MoM? If I can just get the S- or the Z-parameters then it is pretty simple to calculate the correlation coefficient. The correlation coefficient can also be calculated using the far-field.

 

[jian]

Hi, Jone: Antenna Correlation Coefficient is available on the PATTERNVIEW of IE3D EM Simulator. It is based upon radiation pattern. If you don't have IE3D, pelase download a free evalulation from www.zeland.com. I am Jian from Zeland. Best regards,

 

[vfone]

If the distance is greater than the far field region try to use this empirical formula that gives you the approximate coupling between two identical and parallel dipoles (having the same polarization).

Coupling [dB] = 27.6 – 20*LOG (Frequency_MHz) – 20*LOG (Distance_meters)

 

[Azulykit]

It looks like your question could be answered using the Friis transmission formula. It is also know as the one way radar range equation.

Look up "space loss" for a description. You will find it in system design references where "link budgets" are discussed.

Basically power received decreases with the square of the distance from source to receiver. The other terms in the expression relate to the collection aperture which is a function of antenna gain and wavelength.

Just another thought: Using a simulator probably will not work well in this case because the problem size is probably huge relative to the size of the antennas. The demands on the simulation and processor will be large. I assume that the antennas you are studying are very far apart.

 

[Jone]

Maybe I wasn't clear enough: I wasn't talking about one receiving and one transmittning antenna separated by a far distance but about two or more closely placed antennas separated by some fraction of a wavelength or so. Such a configuration is used in MIMO systems. Depending on how the antennas are placed to each other, they will interfere constructively or destructively. A measure of this is the correlation coefficient. Normally you want to place the antennas so that there is no correlation between them. Note also that the correlation coefficient measures the linear dependence, not any other dependence. For example if there is a square relation, then obviously the antenna patterns are dependent on each other but the correlation coefficient is zero.

 

[Azulykit]

Take a look at the array patterns that result form two closely spaced antennas. There are a number of books covering this subject. Kraus and Hansen are good places to start.

If the individual antennas are not too widely spaced you could model the performance and extract patterns and s parameters fairly easily using a 3D filed solver. As the array grows the simulations require more computer horsepower and simplifying assumptions making analysis more challenging.

You might get to where you want to go by placing a time delay after one of your antennas and then performing a vector add of the two signals. That is one way to view what happens in an array with a corporate feed.

I hope this helps.