Although there is a school of thought that because of CSMA/CA it is better to use the same channel rather than a partially overlapping one if complete non-overlapping is not possible. I'm not sure that I subscribe to that view, particularly if one signal is being saturated due to streaming.
The problem is that you cannot see the other multiple sources that use the 2.4GHz allocation, without a spectrum analyser. You can buy a USB wireless receiver, and suitable software, but its not cheap.
Even on my Asus Netbook, with a wireless card that has a promiscuous mode, I can see wireless routers with hidden SSIDs, that are occupying channels. These do not show up on a normal wireless adapter. The spectrum is more crowded than people realise.
Indeed, I think the only real method of selecting a good channel is trial and error!
@licquorice wrote:
Indeed, I think the only real method of selecting a good channel is trial and error!
Exactly, and be prepared to move a single channel up or down, to see if it improves performance.
Use Ethernet or powerline adapters as a first choice, for devices that stay in house, keeping wireless just for those roaming devices where performance is not a major issue.
Things like smart TVs, should only be connected via wireless, as an absolute last resort, as they consume vast amount of bandwidth, leaving every other wireless device stuggling.
@licquorice wrote:Although there is a school of thought that because of CSMA/CA it is better to use the same channel rather than a partially overlapping one if complete non-overlapping is not possible. I'm not sure that I subscribe to that view, particularly if one signal is being saturated due to streaming.
Ive wondered about this.
Its been many a year since i did sideband radio at uni, but that only covered AM/FM tranmission and was before WiFi at home existed 🙂
@Gary_C wrote:
until someone uses 40mhz and spreads even more 🙂
Like the Virgin Media Superhubs
@licquoricewrote:Although there is a school of thought that because of CSMA/CA it is better to use the same channel rather than a partially overlapping one if complete non-overlapping is not possible. I'm not sure that I subscribe to that view, particularly if one signal is being saturated due to streaming.
Looking into it, WiFi uses QAM so the spread is due to the FM part of the modulation spreading around the centre carrier frequency ? So will CSMA look for transmissions only at the carrier frequency before transmitting or the whole used spectrum ?
I would expect it to look at the centre frequency, as modern radio front ends take samples at a specific rate which relates to the centre frequency of the carrier, unlike the old radio receivers that were superhet https://en.wikipedia.org/wiki/Superheterodyne_receiver designs, and had a defined bandwidth.
But its all done by clever mathematics, which I could not start to understand. I was brought up on old radio stuff, using inductors and tuning capacitors
Anyway, getting a bit off topic now.
@Keith_Beddoewrote:I would expect it to look at the centre frequency, as modern radio front ends take samples at a specific rate which relates to the centre frequency of the carrier, unlike the old radio receivers that were superhet https://en.wikipedia.org/wiki/Superheterodyne_receiver designs, and had a defined bandwidth.
But its all done by clever mathematics, which I could not start to understand. I was brought up on old radio stuff, using inductors and tuning capacitors
Anyway, getting a bit off topic now.
Known what you mean. As an instrumentation mechanic originally, all the stuff I worked on used valves not transistors as amps.
so if the carrier sense of CSMA detects the carrier, it might not be able to tell that channel 6 is active when beginning to transmit on channel 7. This seems to be the point some are making in only choosing channels 1,6 &11.