|
Fundamentally, Wi-Fi is an "only-one-thing-at-a-time-can-transmit" technology (although MU-MIMO is mitigating this a little in the downstream direction - with the "right" client mix, AP and physical distribution.)
The more "things" you have, the more data they wish to transmit, the more competition there is for "air time." Estimating whether any particular mix of clients, traffic patterns, latency and so forth is "good enough" is the blackest of the data networking black arts and difficult to model/predict. Welcome to our world.
For any given use case, it might be fine, it might be rubbish, no-one can predict with any certainty and you've little option than to just try it for a while and fix any problems if they become apparent.
Wires are always going to be faster and more reliable, which is why big companies won't be ripping out all the wires in their offices any time soon. Indeed, we often find we have to put more wires in to facilitate Wi-Fi as the ideal location for hanging up the Wi-Fi AP's almost inevitably doesn't have any wires in situ. We often end up building Wi-Fi infrastructure "as well" as the incumbent wired network rather than "instead of."
One of the reasons we put up multiple AP's on big sites as well as or in addition to achieving the geographical coverage, is to the break up the air time competition by creating multiple Wi-Fi cells tuned to different radio frequencies that can transmit concurrently with each other (though there's some limits due to the number of available radio channels.)
Ultimately, it's going to be a value judgement trading off the required speeds, the hassle of installing the wires (versus Wi-Fi) the reliability required, etc. There's no real "right" and "wrong" way to do it. |
|