how to reduce antenna coupling

Hi, I have a question about the antenna coupling.

When two antennas are strongly coupled with each other, and the input signal power ratio to these two antennas changed, what kind of characteristic will be changed correspondingly?

I feel that the radiation pattern, input impedance should vary, but how about the resonant frequency?
Is the resonant frequency of the antenna systems only depends on its geometry, and cannot be changed by the feeding phase and magnitude?

Thanks~

Hi, Jayleung:

For a linear structure (antenna or circuit), if it is one-port, the performance of the structure does not change with the excitaiton magnitude and phase. If it is a 2-port structure, then it is a different story. Changing the relative magnitudes and phases of the 2-ports will change the structure's performance.

I believe the concept of resonance is relative. In the other word, it is hard to say which frequency is the exact resonant frequency because it is dependent upon the loading and how you view it. A typical example is a microstrip antenna. What is its resonant frequency? Some people say it is the Re(Zin) at max. Some people say it is at Im(Zin) = 0 and some people say it is at Min(Gamma). All those points are different even though they are close. Also, for Min(Gamma), it is dependent what Zc you will use. When you have 2-ports, I believe the definition of resonance needs to be clearly defined because it can be some value in a range depending upon how you view it and how the excitations and terminations at the 2-ports are.

Best regards,

Hi, Jian. Thank you very much for your information.

I think most of the people define "resonant frequency" with return loss or VSWR minimum in antenna field.

So I am wondering by adding some tuning feed network for antenna such as L-, Pi-, or T-lump component network, every antenna system can be tuned to match 50 Ohm or other standard terminal at specified frequency point.

However, single antenna element has its best resonant frequency corresponding to its geometry, for example, monopole with lamda/4, rectangle microstrip antenna with lambda/2.

Is it due to the fact that corresponding impedance bandwidth should be deteriorated, or the radiation pattern will be totally changed for tuned matching network added?

jayleung:

An effect that often happens with phase arrays of antennas is something called "blind scan angle" where mutual coupling between elements affects the ability of the array to "see" at certain angles. This is a topic of great interest to electronically steered radar arrays for obvious reasons.

Mutual coupling between multiple driven elements will change the input characteristics of your multi-element antenna array, and can be dependent upon the element-to-element feed (magnitude/phase).

--Max

Hi, Jayleung:

If it is one single antenna, it is easy to find the Min(Gamma) frequency. However, if you have 2 strongly coupled antennas, it may not be so obvious. In such a case, even the definition of Gamma is doubtful. Basically, we have incident and reflected waves (a1, b1) at port 1, and incident wave and reflected wave at port 2. Certainly, we can define gamma1 = b1/a1 and gamma2=b2/a2. However, you may find that Min(gamma1) and Min(gamma2) may not be at the same frequency. Then, what is the definition of resonant frequency of such a case. It might be appropriate to define resonant frequency of 2 strongly coupled antenna as the point where we get minimum reflected power for the whole system of 2 antennas. In this case, we have:

P_ref = 1/2 * ( b1 * b1 + b2 * b2 )
P_inc = 1/2 * ( a1 * a1 + a2 * a2 )

Power_Ratio = P_ref / P_inc

and we need to find the frequency where Power_Ratio reaches the minimum. We also have:

b1 = s11* a1 + s12 * a2
b2 = s21 * a1 + s22 * a2

Apparently, Power_Ratio is related to the s-parameters of the 2-port structure and the magnitude and phase of a1 and a2. For most cases, the Min(Power_Ratio) will neither the point of Min(Gamma1) nor the point of Min(Gamma2) even they can be close.

Best regards,

Jian

Hi Professionals,

Could this power_ratio or resonant frequency be correlated to antenna isolation performance between two application antennas (like for instance, GSM and bluetooth) ? Immediate sense is close distance of two feedings will get strong coupling. BUT will resonant rule of each type of antenna ( like monopole labda/4 as
previous mentioned) be ruined ?? How do we have better isolation if close distance
of feedings is not changeable ?

Thanks and Appreciated !

Hi, Wilson_yu_chen:

When you have 2 antennas closely coupled together, they are inseparable. It is like when 2 circuit elements are closedly coupled, you can not use simulators to simulate each of them and cascade them to get accurate results. Certainly, the strong coupling between 2 antennas will change the resonant frequency of each antenna. How can we avoid the storng coupiling? Certainly, if they are not operating in the same frequency range, it will be the best because you can use filter inside the feed network to filter out the signals. If they are in close frequency ranges, I think you may try to use polarization to isolate them. For example, 2 parallel dipole antennas will have much stronger coupling than two crossed dipole antennas. Regards.

Wilson_yu_chen:

If you have to operate in the same, or nearly the same frequency band, it is possible to try some things that will minimize coupling between the antennas. This can be a bit difficult, but you can sometimes use things like ground screen separations (these can make problems for your antenna pattern, though, so you have to be careful), and for some antennas you can design in ground current chokes (works for some antennas like recessed helix antennas). What you can do to minimize cross coupling depends on the antenna types you are using (and how expensive these measures can make the antenna).

As jian has suggested, it is best to try to keep your antennas in separate bands, and use bandlimited filters near the antenna. Be careful that your filters have low in-band loss, though, for a receiver antenna, or it may affect your in-band noise figure because of the losses.

Good luck,

--Max

mutual coupling is an important item in the array antenna. I think that it is better to used the solutions to reduce the coupling. but this is a question: what is the noneffective value of mutual coupling? -15 dB , -20 dB or higher?

regards

It is certainly nice to reduce the coupling between antennas because designing antenas with coupling included should normally be much harder. However, It is not so simple to reduce the coupling between them when they are close. As every body knows, we would like an antenna to radiate as much as possible. However, when it radiates very well, it also means that it will couple very well to any antenna near it. Normally, low profile antennas are less coupled with the same electrical distance. 3D antennas are coupled stronger. How much is considered to be big? I think it depends upon applications. One thing I can see is that neglecting coupling in simulation will normally over estimate the radiation efficiency. Coupling normally can reduce radiation efficiency. Regards.