What are expected Signal #'s for 5GHz?

[bodean]

What range signal number (use wifi scanner on my mac air) do I want to see for my 5GHz? I think now, when in my family room, it is -67. Is that an acceptable # or should I move stuff to get it lower?

 

[4Leaf]

What range signal number (use wifi scanner on my mac air) do I want to see for my 5GHz? I think now, when in my family room, it is -67. Is that an acceptable # or should I move stuff to get it lower?

It depends on how far you are away from your router?

I'm sure that there is nothing wrong with it though. 5 ghz signal goes wider rather than longer.

 

[TonyH]

What range signal number (use wifi scanner on my mac air) do I want to see for my 5GHz? I think now, when in my family room, it is -67. Is that an acceptable # or should I move stuff to get it lower?

Hi,
The lower the number, the better. 67 is close to be marginal in my experience
based on steady real time HD media streaming environment on our home theater.

 

[stevech]

some devices display in percent. That's not very useful because radio signals attenuate per the inverse square law and a log scale is needed.

some display received signal strength in dBm (decibels relative to one milliwatt). So 0dBm is one mW.

A very strong signal in WiFi is -30dBm or more positive.
As you get to -80dBm the speeds have slowed to minimum and there is little spare margin.

A good number for 2.4 or 5GHz, for a reasonable data rate, is -70 dBm. -60 is better. This is mostly governed by laws of physics. MIMO in some situations, not all, can provide a better data rate at a lower received signal, esp. if both ends have the same form of MIMO. And of course, a good antenna (like 9dBi or better) can improve speeds.

The signal strength in dBm decreases (more negative) as distance increases and/or obstacles become more numerous (like walls, floors vegetation, foliage, terrain).

 

[azazel1024]

Just really depends. Stevetech more or less covered it.

It really depends on what you need/want.

With 11a/g and something like high -60's dBm, you could probably expect to see only 10-15Mbps, but with something like an 11ac router with 2.4GHz 40MHz and 2-stream, you might see 60-100Mbps and with 5GHz and similar signal strengths you might see 150-200Mbps.

Generally you'll be hitting roughly maximum performance in the -40 to -50dBm range (or stronger) and performance will generally start dropping off once you get in to the -50dBm range. -60's is generally when you'll see "mid range" performance and -70s is when performance is going to start really tanking. Depending on the client and the base station some/most can keep on chugging in to at least the low -80dBm range, but performance is going to be very slow.

Only a handful can run up to about -90dBm. Generally around -85dBm is where you'll lose an effective connection.

A lot depends on local noise environment. If you have very low interference (from other wifi networks, or just other sources (like cordless phones, microwave ovens, etc) you might be able to push in to the low -90dBm range depending on the exact sensitivity of the client and base station (there is a minimum floor for all even with zero noise). Other times if interfence is moderately high, you might not be able to sustain a connection even in to the -80dBm range and speeds might start tanking once you hit -50dBm or so.

Generally 2.4GHz and 5GHz are going to perform relatively similarly in terms of where you'll see performance degredation and vaguely by similar magnitudes since they both use similar signaling techniques. Where it gets messier is with 11ac where you might have 5GHz with explicit beamforming, but no support on 2.4GHz, or other technologies that 5GHz might get that 2.4GHz might just not have.

In a an indoor environment though, 5GHz will attenuate a lot faster than 2.4GHz from obstructions, but they'll both generally work to roughly the same signal strength level (5GHz tends to be weaker to start with than 2.4GHz though, due to "stuff").

 

[sfx2000]

some devices display in percent. That's not very useful because radio signals attenuate per the inverse square law and a log scale is needed.

some display received signal strength in dBm (decibels relative to one milliwatt). So 0dBm is one mW.

A very strong signal in WiFi is -30dBm or more positive.
As you get to -80dBm the speeds have slowed to minimum and there is little spare margin.

A good number for 2.4 or 5GHz, for a reasonable data rate, is -70 dBm. -60 is better. This is mostly governed by laws of physics. MIMO in some situations, not all, can provide a better data rate at a lower received signal, esp. if both ends have the same form of MIMO. And of course, a good antenna (like 9dBi or better) can improve speeds.

The signal strength in dBm decreases (more negative) as distance increases and/or obstacles become more numerous (like walls, floors vegetation, foliage, terrain).

11n and 11ac improved the adjacent/co-channel interference specs from +20dB to -9db, mostly as a nod to the popularity of WiFi in both 2.4GHz and 5Ghz...

I would still go with 20dB if one can find a channel here in 2.4GHz - in 5GHz at present, this should not be an issue - there's enough channels, and propagation is less of an issue here for neighbouring AP's...

 

[stevech]

I would still go with 20dB if one can find a channel here in 2.4GHz - in 5GHz

unclear what "20dB" relates to, s.v.p.

 

[htismaqe]

unclear what "20dB" relates to, s.v.p.

I think he meant to say "20Mhz".

 

[sfx2000]

unclear what "20dB" relates to, s.v.p.

relative to the co-channel/adjacent channel transmitter (e.g. other AP).

One of the upsides of more recent 11n/11ac clients - they do have better baseband/RF blocks in them

 

[stevech]

relative to the co-channel/adjacent channel transmitter (e.g. other AP).

To me, co-channel is frequency reuse, e.g., an access device is on channel n and another is on the same channel, and their coverage overlaps. This is a timing problem. With 802.11's CSMA/CA, it should usually self-resolve so there are rare collisions, though the two share the channel's traffic capacity. Adjacent channel (like 1, 6, and 11 are the 20MHz channels), is a different issue, since CSMA/CA has to use energy detection rather than preamble detection. Adjacent as in ch 6 with 7 is of course worse, in that the channel power is much higher.
One of the upsides of more recent 11n/11ac clients - they do have better baseband/RF blocks in them [/QUOTE] "RF Blocks", meaning channel-wide filters? At the baseband? Huh? Some of the newer modulation modes do reduce the occupied bandwidth, e.g., from 20 to 18MHz, but the front-end of WiFi receivers is wide open to the entire band. Tunable sharp-skirted filters are way too costly for consumer gear.