10 Huge Wi-Fi Antenna Mistakes

|
10 Huge Wi-Fi Antenna Mistakes

 

Like I outlined in my previous blog, antennas are undervalued and poorly understood elements in WLAN networks. Even Apple is prone to making huge mistakes if you remember the famous “Antennagate” scandal that came about as a result of poor antenna placement and design on their new iPhone.   And as Steve Jobs pointed out, antenna problems are not unique to iPhone or Apple.  Many times, antennas are the root of Wi-Fi performance problems. But if you understand RF, then there are common sense things we can follow and avoid.   Let’s consider 10 common antenna mistakes and their impact on the end user Wi-Fi experience.

Poor antenna placement

Mistake 1
• Access points with integrated omni-directional antennas (where the max gain is sideways) are placed among or above metal air ducts, grids and large lamp reflectors.
• Outcome: A lot of reflections and refractions cause multiple copies of the signal to arrive at receivers, thereby dropping the data rates. As the radio waves reflect in random directions, the signal strength suffers at both ends.

Mistake 2
• Access point integrated dipole antennas stick partially through a small hole in a metal ceiling panel, AP cover or metal grid.
• Outcome: The antenna element has detuned and part of the energy is now directed upwards. Antennas protruding halfway in this manner create a total mess.

Mistake 3
• An antenna element is placed on top of a ceiling panel made of unknown material with ɛr > 1.
• Outcome: Antenna is detuned to another frequency, the radiation pattern and impedance matching are altered at the target operating frequency. Antenna gain becomes attenuation.

Mistake 4
• Access points with standard omni-directional antennas are placed on a high ceilings — 4-5 meters high (13-16 feet.)
• Outcome: Most of the energy is directed sideways towards other access points causing co-channel interference. Also, antenna gain shoots over the target area and data rates are reduced.

Mistake 5
• Directional, high gain omni-directional antennas are used and placed on a high ceiling.
• Outcome: These antennas have narrow beams pointing sideways, therefore, significant overshooting and even more co-channel interference takes place.

We’ve put together a comprehensive guide on antenna placement and we invite you to download it by clicking below.

Get Antenna Placement Guide

Mistake 6
• Access points with omni-directional antennas are placed next to a thick wall.
• Outcome: Half of the antenna’s gain is lost as the antenna tries to pick up signals from the wall, which is an unused direction.

Mistake 7
• Access points with sideways gain patterns and designed for ceiling mounting are installed on a wall.
• Outcome: The gain goes up and down, amplifying interference and decreasing signal levels.

Mistake 8
• Antennas are painted with a metallic paint.
• Outcome: Signal attenuation due to the conductive surface.

Mistake 9
• End user terminals (such as laptops) are placed in metal enclosures, like carts.
• Outcome: the proximity of the conductive material causes antenna detuning and the energy cannot radiate outside the box. A very large attenuation in signal levels is the result (>10 dB).

Mistake 10
• Often it is forgotten in network planning that VoIP terminals are placed against the human head with a hand covering the other side.
• Outcome: Site surveys with laptop card and standard limits give completely different results.

Despite the challenge, there are some rules of thumb that help with getting good performance out of antennas:

  • Larger antennas have typically better performance than smaller ones. Gain/directivity is not available from a very small surface area.
  • Know and understand your antenna radiation patterns. Point antennas gain towards users.
  • Make antennas visible to end users, do not hide them above the ceiling grid.
  • Do not place anything conductive near antennas. Allow 3 feet of distance, preferably more.
  • Use good quality external antennas, as the integrated antennas often have been compromised for design, though there are some exceptions.
  • Simple dipole antennas or four-six leg spider-like APs are rarely an optimal selection.
  • If you use omni-directional antennas, look for ones that have down tilt and attenuate upwards signals clearly. For example, max antenna gain does not go straight sideways but travels partially down.
  • In high ceilings, use relatively high gain(6-9 dB) patch antennas directed downwards.
  • Use antennas that support all MIMO inputs/outputs available from an access point.
  • Single frequency antenna elements often perform better than dual tuned elements.
  • When available and supported, always use diversity antenna at the AP. Remember to enable it from the radio configuration settings.
  • Allow 5 dB for additional hand/head attenuation for VoIP terminals.
  • Remember to adjust AP output power in relation to antenna gain.
  • Place APs with down tilt antennas horizontally to the ceiling, not vertically to a wall.

Finally, it is fun and entertaining to visit public places and observe the “innovations” in WLAN implementations, though on a rare occasion, there are some good ones as well.

One funny story was where a prison inmate was using a Playstation WLAN to communicate with others outside the prison. What made this innovation entertaining was the fact that he was using a large frying pan lid as a reflector to increase antenna gain and achieve the range needed for connectivity.

With all the hype, are you confused by 802.11ac? Not sure what to believe? Click below if you would like a no spin, honest take on what to expect.

802.11ac Migration Guide

23 thoughts on "10 Huge Wi-Fi Antenna Mistakes"

  1. Tom E. Christensen says:

    Very good info! Thanks 🙂

  2. Gary says:

    Your are quite right, antennas are the most undervalued poorly understood and influential part of any WLAN.
    I use a number of specialists antennas for different scenarios and wish I had a greater choice.
    Oh yes, frying pans are a truly great invention, and so it seems are their lids!

  3. Jan Kruys says:

    Good stuff from 7Signal – as usual. These considerations – and many others related to RF and signal propagation – should be part of every Wi-Fi Engneerig course but I doubt if that’s the case.

  4. Gaurav Narain Saxena says:

    Thanks for publishing an article that makes my jaw hit the floor with astonishment. Also double thanks for emphasizing in an indirect way that people should really ‘know’ their dwellings for maximum success with their wifi based assets.

    1. Veli-Pekka Ketonen says:

      Tom, Gary, Jan and Gaurav, thanks for your comments. Very much appreciated!

      One interesting topic related to antennas, more exactly to power splitters, came to my mind. In some specific cases, for example large refrigerators in fresh food warehouses, it may be useful to split one AP/coax RF feeder to two antennas and avoid putting a dedicated AP inside all of them. Not exactly necessary to have 802.11ac 3X3 MIMO AP feeding with >1Gbit/s to that one handheld bar code reader used inside the metal container few times/day.

      Take typical RF power splitter operating up to 6 GHz frequencies. Splitter has one incoming coax cable and two going out. If there are coax cables at both outputs, it obvious that power is split in in two parts. So at both outputs there is about 50% of the incoming power in milliWatts minus some implementation loss. So total splitter loss is about 3.5-4 dB. Remember the rule that +/-3 dB means always double or half the power in Watts.

      Now the mystery: It’s interesting what happens when only one splitter output is connected to a coax cable and the other output is disconnected/empty. Since RF energy can’t just disappear, first thought in this case is that all power goes to the connected output. Interestingly this is not the case. Output power in that one connected output is in practice the same independetly if both outputs are connected or only one…but hey, with one output only where does half of the RF energy disappear! 🙂

      There are some great RF articles in free publications like:

      Microwaves and RF
      http://mwrf.com/

      Microwave Journal
      http://www.microwavejournal.com/

      Microwave Engineering Europe
      http://www.microwave-eetimes.com/

      1. Primož Marinšek says:

        All are great examples and I especially like the last one about the phones being placed next to a head.

        But to comment No. 6. Although it is a mistake to place an AP next to a wall or a corner many times you aren’t allowed to put a directional antenna somewhere due to aesthetics or other reasons and the only way to control RF propagation and avoiding CCI is by placing it there.

  5. Lee Badman says:

    Great article, and factual for sure. I have done antenna work in a number of realms for my entire adult life, and find them fascinating to build, select, and use (ham radio geek on the side). The one thing I kind of get burned out on though is the “though shalt” guidance that calls out the sub-optimal without ever giving a nod to the reality that we often have no choice but to violate best practices. There are spaces where are appearance trumps all, where there are high, finished ceilings, and never in a million years would you get away with anything less than an imperfect deployment for the sake of aesthetics (like when external antennas are forbidden, or where you HAVE to mount on walls at silly heights). It’s easy to point at specific deployments and say “that would perform better if it had been done differently!” but for many of us, often we have no choice but to stray from best practices. Complex environments mean becoming an artist in making due when you can’t do what you know SHOULD be done for a number of reasons. The key is learning how to to make the most out of the circumstances you have to work with when you can’t do it 100% “right”.

  6. BruceR says:

    If you have an office deployment where all the suspended ceiling is made of metal panels, is there some way to mitigate the effect.

    Would just replacing the panels that the APs are fixed to help.

    1. Veli-Pekka Ketonen says:

      Do you mean that APs are already below the metal panel or are APs are above and dipole antennas come through the panel?

  7. Robert Fauvelle says:

    Great article and useful information. Thank you.

  8. Primož Marinšek says:

    All are great examples and I especially like the last one about the phones being placed next to a head.

    But to comment No. 6. Although it is a mistake to place an AP next to a wall or a corner many times you aren’t allowed to put a directional antenna somewhere due to aesthetics or other reasons and the only way to control RF propagation and avoiding CCI is by placing it there.

  9. T. Eternus says:

    The article continues to make a strong case for WLAN surveys and whenever possible practical testing. I don’t know why more people don’t do some kind of testing before deployment.

  10. Dennis says:

    I own a Linksys ea6900 wireless router and I am shopping for new antennas. My question is: For best results/speeds should I use 3 antennas with different dPi ratings? Or should I use 3 matching antennas? My property is a decent size ( 10 acres ) and I’d like to have total coverage.

  11. Gareth Price says:

    ref Mistake 6.

    Indeed point taken, but a number of manufacturers provide wall mounted omni-directional units and where ceiling units are not possible where else are you going to provide above desktop positioning? A single wall mounted directional antenna is not the best solution for large offices? The best positioning for a dual omni-directional unit would be in the centre of the room on a 10ft tall pole right? Not happening in a plush modern office.

  12. ip info says:

    Could you give some simple guidance on simple plane reflectors for wifi. These could be useful if the router has to be placed near an external wall, say. In particular, how close should a reflector be ( given the 2.4Ghz range has a wavelength of 12.5cm). In this case, I guess a reflector at 6-7 cm would cause major problems and one at about 1 cm would be quite helpful. Comments please!

  13. ip info says:

    Your publications have been very helpful in helping others understand the properties of rf and it transmission. Being a radio amateur for year and working with radar as well as avionics, these questions are helpful to explain to others the principals involved.

    Thanks
    Barry Robinson

  14. ip info says:

    You are correct 100% . All wifi antennas are made up of rubber duct outside layer and inner layer with highly flexible soft conductor metal. It is very sofisticated and delicate too due to this signal capacity of radiation of signal is dipressing and generating comb noise between client device and roughter. To reduce these kind of radiating hall effects. I strongly recommend to use pure crome metal antenna because of its properties and chemical structure to radiate SUHF upper band frequencies in narrow mode.

  15. Jon Sullivan says:

    I found what appeared to be the exact same antenna that is made by well known antenna manufacturer but for 25% less. These are offered by a well-known, well respected wireless distributor so they must be OK right?

    When we received the first sample antennas our RF techs confirmed the gain and return loss were within the specification.
    Unfortunately we could not verify where they had been manufactured or if they had been electrically checked. The “certificate of compliance” was written in Chinese and the specification sheet seemed to be copied (verbatim) from our original supplier’s specification. We removed the radomes and found the antenna elements (radiator, feed lines and ground) had been made from materials that were not protected from corrosion.
    I am sure that over time, these would corrode and cause the return loss to shift and the efficiency gain to degrade. BE CAREFUL WHEN BUYING LOW COST ANTENNAS

  16. gene says:

    ADDS… and instructions never mention this so how do we know?

    If you use omni-directional antennas, look for ones that have down tilt and attenuate upwards signals clearly. For example, max antenna gain does not go straight sideways but travels partially down.

  17. Aaron C. says:

    Good info but one thing you missed. Often improper alignment is a major culprate for improper antenna placement.

  18. putlocker123 says:

    The chamber afforded us a controlled environment to make sure we’d be able to see any “signal” or difference in the KPIs produced by the 7Signal system with and without LTE.

Leave a Reply

Your email address will not be published. Required fields are marked *