Why use semi-directional antennas, beamforming, higher transmit power, etc. to increase AP range if you cannot increase the transmit power of the client STA as well? Wouldn't these methods of increasing the range of the AP be bottlenecked by the limited range of the client anyway?
The short answer is the client STA is almost always the limiting factor on AP range, and that's a good concept to understand. That being said;
A directional antenna would allow you to increase usable coverage distance (not necessarily coverage area) even without raising the Tx power on the client STA. The AP's energy would be more focused at the client and the directional antenna would allow the AP to better hear signals from the (presumed) omni antenna in the client STA.
Beamforming is more about improving link quality than actual range, although greater usable range often is a result.
Higher Tx power will only be effective if you have it set far too low to begin with. Tx power that is too high is exactly what causes near-far issues where the client can Rx the AP traffic but not vis-versa.
Remember, our goal as wireless engineers is to achieve the best possible user experience. In every single one of my projects, that is achieved with a laundry list of small improvements, and all of those improvements are site/customer specific.
It may not affect your question, but I have found (real) Sensitivity is a much bigger factor than output power for range.
WLAN design typically starts with figuring out the S.O.M. System Operating Margin. SOM is typically based on the worst STA which typically means the STA with the worst “receive sensitivity”. Typically todays WLAN’s are being designed for maximum capacity.
When was the last time you surveyed for range?
If your using say a Cisco 1262(2.4GHz) with 5db Omni’s @13dBm say your output power is 15dBm (-3 for connectors)
Free space path loss @50 feet is 64dB
If you’re looking for -67 for the STA then your SOM is 18dB, or quality WLAN coverage.
Hope this helps?
Yes Agreed! A Link between AP & STA is always UPLINK limited & most of the survey engineers overlook this part.
There's catch here. When you talked about all these features, including Power, Beamforming etc, you didn't talk about EIRP/Antenna part.
Someone just pointed out that AP Sensitivity is more important criteria, and I agree with him. With Carrier Wi-Fi, most of the time you would find end customers always evaluate two models based on Rx Sensitivity & EIRP. Rest doesn't matter most of the time (my experience).
So what you do? Wrong approach would be to increase AP power, but increase EIRP by putting high gain directional antenna to improve on UPLINK limitation.
Hope this helps.
Lots of replies hitting on this idea but I'd like to state it specifically - antennas provide bi-directional gain; that is, to both the AP and STA. Though the gain will not be equal in both directions (always more from the transmit side at the antenna), the higher gain of directional antennas will affect both the AP and the STA.
Can you provide a cite for "Though the gain will not be equal in both directions..."?
My understanding of the "Theorem of Electromagnetic Reciprocity" indicates that the receive gain and the transmit gain should be the same. I understand the exceptions, but they mostly don't apply to the antennas commonly used in Wi-Fi (and similar) systems.
I've reread the post and my response and apologize for having wandered into old habits regarding directional antennas, which I often tend to translate as "parabolic". Parabolic antennas with subreflectors exhibit more gain in the transmit direction than in the receive direction as a function of the focus of the beam at the local source vs. the distant end. While this may not be true in other antenna types, and is not often delineated in theoretical discussions, it is a known (measured) characteristic of parabolics (which, again, I mistakenly referenced). Antenna datasheets (such as http://www.gatr.com/products/2-4-antenna-system) typically reflect this characteristic.
The difference between send and receive gain seems to be greatest in the C-band. Is this a wide-band antenna/wavelength design compromise ?
If so, could you design a narrow bandwidth antenna with equally good send/receive performance ?
As you can see from the listed specifications, the lower frequency/longer wavelength results in greater differentiation of gains for tx and rx. So, yes, I believe that it is a part of the wide-band antenna/wavelength design compromise that you mentioned. However, I'd suggest that beamwidth is a factor, as well. Although I did not intend to get this far in our discussion/research, I would suspect that you would see a convergence of gains as you narrow bandwidth, but do not believe you would have the same result by narrowing beamwidth.