Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Antenna Gain

  1. Apr 13, 2009 #1
    Hi all,

    Does the antenna gain depend on the angle of the emerging ray? or It can be a single number for all ray emerging in all directions?

    I know the gain is the amount of power an antenna has relative to an isotropic antenna. Sometimes it means the maximum power an antenna can have relative to the isotropic antenna.

    But if it is a directional antenna, each direction would have a different gain. Can we use the maximum gain in equations like:
    Pr = Pt Gt Gr / 4 PI r^2 and Effective Area = Gain * Effective Area of an isotropic antenna?

    I'm using an antenna of frequency 2.4 GHz
    Last edited: Apr 13, 2009
  2. jcsd
  3. Apr 13, 2009 #2


    User Avatar

    Staff: Mentor

    Yes. For a directional antenna, the gain is a function of direction. In this case, when people talk about "the" gain, they're usually referring to the maximum gain.

    No, you have to use the gain appropriate for the direction of orientation for each antenna. If the transmitting and receiving antennas are oriented so their maximum-gain directions are aligned, then you can use the maximum gain for both of them.
  4. Apr 13, 2009 #3
    In access points data sheets they just mention the gain of the antenna. How can I conclude the rest of the radiation pattern? How can I make use of that gain?

    Is there also a gain for wireless cardbus adapters?
  5. Apr 14, 2009 #4


    User Avatar
    Science Advisor
    Gold Member

    Most antennas will have a similar radiation patterns. For example, small antennas all pretty much have the same radiation pattern regardless of their shape (Steven Best has done a bunch of papers on this). Most patch antennas will be rather similar. It usually isn't until you start using a directional antenna or arrays that you get vastly differing radiation patterns.

    A wireless cardbus probably uses something like a monopole or patch antenna (you can tell from looking at it probably) and either way you can probably just estimate it roughly as the pattern for a half-wave dipole.
  6. Jun 22, 2009 #5
    This is rather an old thread but I had to ask this question.

    I lack any experience in antennas so what kind of antenna does this http://www.cisco.com/en/US/prod/collateral/wireless/ps5678/ps6087/product_data_sheet0900aecd806351fc.html" [Broken] has?

    and what's its gain radiation pattern? (popular approximation would be fine)

    Last edited by a moderator: May 4, 2017
  7. Jun 22, 2009 #6


    User Avatar
    Science Advisor
    Gold Member

    The literature says that it is an omnidirectional antenna. The actual antenna can vary, I have seen some interesting wideband antennas that are used in these kind of appliances and they can look a bit esoteric if they are not wire antennas and for wideband applications (like cell phone reception inside buildings). The data sheet gives the gain as 3 dBi so I would just model it as a simple omnidirectional pattern (dipole antenna) that is scaled accordingly. I don't know if the antenna would have a ground plane or not but that would be another change to the radiation pattern if you want that assumption.

    EDIT: Most of the time the radiation pattern is designed for two very different goals, either highly directed or omnidirectional. If you are making a radar or a repeater, you want a highly directional beam. If you want a general access receiver and transmitter, like for cell phones, radio (terrestial or satelite) or wireless LAN, you want an omnidirectional pattern. The actual patterns will differ whether or not they make use of a ground plane for image theory or due to the dependence on frequency over the desired bandwidth. You really are not going to know the unique characteristics of an omnidirectional antenna unless you simulate or measure the actual antenna.
  8. Jun 22, 2009 #7
    I just want to make sure of something:

    having 3 dbi antenna gain means that the directivity is 2. If I'm using a half wave dipole antenna, then I will multiply the normalized radiation pattern by 2.

    is this correct?
  9. Jun 22, 2009 #8


    User Avatar
    Science Advisor
    Gold Member

    No, gain is the directivity scaled by the antenna's efficiency. Usually we take the efficiency as the radiation efficiency. Radiation efficiency is the ratio between the amount of power radiated versus the amount of power inputted to the antenna. Small antennas and wire dipole antennas working at non-resonant frequencies are examples of low radiation efficiency antennas but their directivities will all be similar. You do not want to use directivity, you want to use gain for your radiation patterns as the gain will take into account more attributes of the antenna. Other attributes that you could use would also be scaling the gain by the matching efficiency but this is dependent upon the network that the antenna is hooked up to and so it usually is not included in simulations (except for direct comparison to measurements or exact simulations of a specific system). What you should do is convert dBi into dimensionless units using 1 as your reference level for the isotropic antenna. Then you simply multiply the gain pattern of the isotropic radiator with the gain pattern of you desired antenna (omnidirectional, what is that like sin \theta / \theta ?) and then scale it by the gain. In this case, 3 dBi would be 2.0.

    So in other words, just replace the word "directivity" in your statement with gain and it is correct.
  10. Jun 22, 2009 #9
    so what's you are saying is that i should consider 1 as the reference gain for my reference isotropic antenna and then divide 3/1 (in that way i would have unitless gain). Then i convert the dB to normal scale by 10 ^ x/10. Finally i multiply gain of my antenna by the result right?
  11. Jun 23, 2009 #10


    User Avatar
    Science Advisor
    Gold Member

    The decibel scale they are using is in log_10, technically I think normalization here is 10 log_10 and it is referenced to the isotropic antenna's gain (hence why we call it dBi). Thus, if we normalize our iostropic radiator to a uniform gain of 1, then 3 dBi would be approximately a gain of 2.

    But to correct your statement, you would not divide 3/1. This is already in decibels and in reference to the isotropic radiator. So you would want to add the gain of the isotropic radiator in decibels. Thus: Desired Gain = Gain in dBi + Gain of Isotropic Radiator = 3 dBi + 10 \log_{10} 1 = 3 dBi + 0 dB = 3 dB. And from here you convert to dimensionless units which will be ~2. This is your gain, so you would multiply your normalized radiation pattern (normalized to however you normalized the Isotropic Radiator to) to get the gain of your antenna (as opposed to "multiply gain of my antenna by the result right").

    A note on dBi, sometimes the gain will be referenced to the gain of a dipole antenna. That is done more rarely but I have seen it once or twice. It should just be pointed out that you understand dBi with the primary point here being that decibels are calculated in reference to something.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Antenna Gain
  1. Cellular antenna (Replies: 3)

  2. The human antenna (Replies: 5)

  3. Antenna length (Replies: 2)