On Dec 18, 2019, at 8:25 AM, Pavel Nikulin <pavel@xxxxxxxxxxxx> wrote: > Second to that, a minute disturbance like a person passing in front of > an AP, or something electrically conductive being put near the antenna > may also easily add or subtract few decibels from the signal level > quickly. I think there should be at least some moving average used for > the roaming decision, or better a more advanced filtering method. The following story is relevant. When Microsoft first did their WiFi implementation, an engineer was put in charge of the driver. He believe that the system should switch to another AP, as soon as it found a stronger signal. The architect involved disagreed. After some technical discussion, the architect finally showed up at the engineers desk. He asked the engineer to switch his internet connection to WiFi only. And then stood between the computer and the AP in use. When the system switched to another AP, the architect moved to block that one, too. After a few repetitions of this process, the engineer agreed that it wasn't a good idea to immediately switch to a stronger signal. The algorithm they implemented was to stick with one AP until the signal got too bad, and then switch to the strongest signal available. It means that connections can be poor for a while. But the important thing is that there is still a connection. Just a cute story to show the real-world impact of these decisions. :) Alan DeKok. _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
