What are the "BSSID Basic Rates" I see when using "netsh wlan show all" command?

[jegesq]

Using a command prompt in Windows, if I do a "netsh wlan show all" command, returns (among a lot of other info), the following:

BSSID 1: ASUS_5G

Network type: Infrastructure
Authentication: WPA2-Personal
Encryption: CCMP
BSSID 1: [Mac Address here]
Signal: 99%
Radio Type: 802.11ac
Channel: 161
Basic Rates <Mbps>: 12 24 58.5
Other rates <Mbps>: 18 36 48 54 526.5

What are the "Basic Rates" and the "Other Rates" in this context? Are they theoretical or real transfer/receive rates for this particular BSSID (i.e, 5.2ghz channel 161 on this radio with my current wireless "ac" adapter? And if so, why is the top end "rate" 526.5?

Any info would be appreciated.

 

[htismaqe]

526.5 is the maximum theoretical data rate for 2-stream, 80Mhz channels, MCS 6.

https://en.wikipedia.org/wiki/IEEE_802.11ac#Theoretical

The "basic rates" and "other rates" fields are SUPPOSED to show rates included in the standard and all supported rates, respectively. I've never actually used them, so I'm not sure why yours only show those data rates.

 

[jegesq]

Thanks....um, I think.

After reading the Wikipedia article and your answer, I see the chart that shows "Theoretical throughput for single Spatial Stream (in Mbit/s)" and if I look down the MSC column to the row "6" line, I look across to find the column marked "80mhz" and then look at "800 ns GI" and see that the maximum theoretical throughput for a single stream is "263.3". Am I reading this correctly? And what is "800 ns GI"?

So if my RT-AC66U and my Intel 7260ac adapter are both connected using two spatial streams on 80mhz channel width, my theoretical throughput is thus doubled, and it's 526.5. Am I reading this correctly?

If so, then how and why does this theoretical throughput rate in Mbit/s differ from the chart showing "Advertised" rates (which I'm guessing means the highest theoretical rates possible that are advertised --as in marketing advertising--for each class of AC device shown in the chart). Or does it mean that the router is "advertising" (transmitting) this rate as being available to client devices? I'm unclear I guess what "advertised" means in this context.

For example, if I locate my laptop with the 7260AC adapter right next to m AC66U (say within 1 or two feet of the router) the Phy Layer connection speed is listed as 867? I'm guessing that because my 7260 adpter is only an AC1200 device that it will list this rate as the "maximum" for throughput. But I thought from the chart above that the theoretical maximum is 526.5? So which is it? Or am I comparing apples and oranges?

Any additional insights you or anyone else can offer will be appreciated.

 

[htismaqe]

After reading the Wikipedia article and your answer, I see the chart that shows "Theoretical throughput for single Spatial Stream (in Mbit/s)" and if I look down the MSC column to the row "6" line, I look across to find the column marked "80mhz" and then look at "800 ns GI" and see that the maximum theoretical throughput for a single stream is "263.3". Am I reading this correctly?

Yep.

And what is "800 ns GI"?

800 nanosecond Guard Interval. It's rather complicated but it has to do with signal reflection and the delay introduced to allow for signals to dissipate.

http://www.connect802.com/802.11acad-Page04.htm

Someone like SFX2000 or Tim would probably better-qualified to give a complete answer but the bottom line is that 800ns GI is more common.

So if my RT-AC66U and my Intel 7260ac adapter are both connected using two spatial streams on 80mhz channel width, my theoretical throughput is thus doubled, and it's 526.5. Am I reading this correctly?

Yep.

If so, then how and why does this theoretical throughput rate in Mbit/s differ from the chart showing "Advertised" rates (which I'm guessing means the highest theoretical rates possible that are advertised --as in marketing advertising--for each class of AC device shown in the chart). Or does it mean that the router is "advertising" (transmitting) this rate as being available to client devices? I'm unclear I guess what "advertised" means in this context.

The advertised data rate is the aggregate, maximum throughput of the device - the maximum 2.4Ghz rate and the maximum 5Ghz rate added together. In this case, advertising = marketing. It's not a technical term.

For example, if I locate my laptop with the 7260AC adapter right next to m AC66U (say within 1 or two feet of the router) the Phy Layer connection speed is listed as 867? I'm guessing that because my 7260 adpter is only an AC1200 device that it will list this rate as the "maximum" for throughput. But I thought from the chart above that the theoretical maximum is 526.5? So which is it? Or am I comparing apples and oranges?

It depends on the modulation and coding scheme (MCS) value. Your devices would have to support 256-QAM (MCS 8/9) to get 867 with just 2 streams and it would have to negotiate 400 ns GI rather than 800. Again, this is probably best left to someone that actually works on these devices at the silicon level like SFX2000 or RMerlin.

 

[jegesq]

Thanks so much. I think I now understand this better. You've been very helpful. I need to read more on this though because I want to get a better grip on the QAM aspect. I have seen, for example, in my cable modem's GUI (it's an Arris Motorola SB 6183 with 16x4 bonding) that all 16 of the channels that are bonded by my cable company to transmit internet to me are QAM256 channels, but I have no idea how that relates to what is ultimately transmitted to the router from the modem (if it relates to it at all after the modem gets done "modulating" and "demodulating"....lol).

But you've been very helpful. Thanks again and I hope that SFX2000, Tim Higgins or Merlin will add whatever they can to enhance understanding of this.

Again, thanks so much. I need to read a lot more.

 

[azazel1024]

400ns (short guard interval) is standard for 11n and 11ac. It is not standard for 11a/b/g. This is the nit that most routers have that is listed as "disable short guard inverval".

In effect, it is more about interference than it is about signal dissipation. When you have a lot of different devices talking, a short time between packets can mean additional interference, so a longer guard interval can actually increase throughput, but generally only in high interference environments.

In low interference environments, a short guard interval is very much desired (IIRC, Short GI generally results in around a 20% throughput increase. In theory).

 

[htismaqe]

This is the nit that most routers have that is listed as "disable short guard inverval".

I have yet to use a consumer router stock firmware that allows the setting of the guard interval. I know for certain the neither Netgear nor Linksys allow it to be set - it's automatically negotiated. I'm not even sure it's exposed in DD-WRT.

 

[azazel1024]

I have yet to use a consumer router stock firmware that allows the setting of the guard interval. I know for certain the neither Netgear nor Linksys allow it to be set - it's automatically negotiated. I'm not even sure it's exposed in DD-WRT.

All TP-link routers I've ever seen or used have the option to "disable short guard interval". Several of the, admittedly older, Netgear routers I've used also had this feature.

 

[htismaqe]

All TP-link routers I've ever seen or used have the option to "disable short guard interval".

Interesting.

Several of the, admittedly older, Netgear routers I've used also had this feature.

My first Netgear routers were WGR614's and the WNDR3700, starting about 5 years ago, and I've never seen it.

 

[jegesq]

Ok, now you've piqued my interest. When and why would you want to adjust your Guard Interval?

Is the setting on TP-Link routers you've seen to "disable short guard interval" the functional equivalent of "enable long guard interval?"

BTW, GI doesn't seem to be among he settings exposed in the RT-AC66U. Is it exposed anywhere in any other ASUS routers?

 

[azazel1024]

Interesting.



My first Netgear routers were WGR614's and the WNDR3700, starting about 5 years ago, and I've never seen it.

My WNDR3500L v1 had it.

 

[azazel1024]

Ok, now you've piqued my interest. When and why would you want to adjust your Guard Interval?

Is the setting on TP-Link routers you've seen to "disable short guard interval" the functional equivalent of "enable long guard interval?"

BTW, GI doesn't seem to be among he settings exposed in the RT-AC66U. Is it exposed anywhere in any other ASUS routers?

Probably it is functionally the same.

The reason to do it in high interference settings it can actually improve throughput (sometimes). With longer pauses you may get fewer times that your transmissions get stepped on, resulting in fewer retransmits. The length of the pause is significantly shorter than the transmission itself, so if you have, say, half the retransmits, even though you have a longer pause between, it can result in higher actual through put.

In some brief testing I haven't seen a scenario where it improved throughput (only a loss of 10-20%). That said, my usage is extremely low interference as I have no real neighboring networks.

 

[htismaqe]

My WNDR3500L v1 had it.

Are you sure you're not thinking of the preamble setting?

EDIT: Nevermind, I did find a reference to the setting in the user manual for the WN802. Of course, that router was released in 2006.

 

[azazel1024]

Are you sure you're not thinking of the preamble setting?

EDIT: Nevermind, I did find a reference to the setting in the user manual for the WN802. Of course, that router was released in 2006.

It had both long and short preamble as well as the ability to disable short GI interval. It actually had quite a few more nits than my TP-Link routers do.

 

[htismaqe]

It had both long and short preamble as well as the ability to disable short GI interval. It actually had quite a few more nits than my TP-Link routers do.

I think that's probably common across a lot of vendors, and a lot of tech products, not just routers.

Widespread acceptance/adoption leads to "user friendly" interfaces.

 

[azazel1024]

Yeah. Getting to the point where soon all we'll have on routers is "on/off" switches and maybe if we are lucky "router/AP mode" switches.

 

[jegesq]

Are you sure you're not thinking of the preamble setting?. . .

Speaking of preamble and settings, there are three different settings in my RT-
AC66U's GUI under "Wireless>Professional" for both 5.2ghz and 2.4ghz bands: "Long" (which appears by default), "Short", and "Auto."

First, does the "Short" or "Long" settings have any effect at all when using 5ghz channels? I ask because "Long" as I understand it, is used to stabilize and improve compatibility between the router and older 802.11b devices. And since devices using 802.11b are incapable of utilizing the 5ghz band of channels, and can use only 2.4ghz, does it really make any difference whether one's 5ghz preamble settings are "Long", "Short" or anything else (i.e., "Auto")?

As I read the DDWRT wiki's entry for Preamble, the entire concept of newer wireless routers even having a "Long" preamble setting is simply to insure backwards compatibility with 802.11b devices, and it's just not necessary at all if you have no such devices to use a "Long" setting. Is this a correct understanding, or is there more to it when it comes to Preamble settings?

Second, I note that the Tomato page describing "Preamble" settings has a separate section devoted to "802.11n Preamble Settings" which read as follows:

"802.11n Preamble
"Like 802.11g, 802.11n transmits a signal that can't be decoded by devices built to an earlier standard. However, 802.11n operating in "mixed" mode transmits a radio preamble and signal field that can be decoded by 802.11a and 802.11g radios. 802.11n Wi-Fi networks have an optional "greenfield" mode that improves efficiency by eliminating support for 802.11a/b/g devices."

So, if I have no b or g devices at all in my environment, should I just set preamble to "Short" (or "Auto" in case someone with an older g device happens to visit my home)?

Third, assuming it does make a difference for the 5ghz band settings which setting I use for Preamble, would using the "Auto" setting cause the wireless router to broadcast either a Long or Short preamble depending on what sort of device is attempting to connect? If not, then what else would "Auto" mean in this context?

Or is the "Auto" setting more akin to the "Greenfield" mode, in which support for a/b/g devices is just eliminated entirely?

Any thoughts or clarifications?

 

[azazel1024]

Set it to short unless you own any a/b/g devices.

Short breaks 11b. Other stuff is broken with Greenfield mode (a/b/g).

Auto will detect if there are 11b devices attempting to connect (because the beacon signal is still sent out at 11b rates and readable by 11b clients, that is unless you have the option to change the beaconing rates, which IS possible, but few routers have the option) and it will fall back to long preamble if such a device attempts to connect.

Frankly I turn off as much legacy stuff on routers as I possibly can. The only 11g device I have, has its wireless turned off (wireless network printer, but it is wired in to my network). Everything else is 11n or 11ac and frankly if you bring a phone, tablet or laptop over that is old enough that you only have 11g in it, you should be ashamed and I don't want you on my network.

 

[jegesq]

Thanks for the info, and for clarifying the "auto" setting for preamble. And while the AC66U's GUI does appear to have the option to allow for changing the beacon interval rates, I would rather not mess with that at all (since I have no idea what rates would turn off the b signal and what would leave it on), so as you suggest, it just seems much easier and simpler to turn off as many of the settings that enable legacy devices (e.g., b) in the first instance.

I got a chuckle out of this: "...and frankly if you bring a phone, tablet or laptop over that is old enough that you only have 11g in it, you should be ashamed and I don't want you on my network."

Me too.

 

[azazel1024]

I believe, anything over 1Mbps beacon will break 11b (it won't even see it). I THINK 11g should be compatible with anything up to and through 11Mbps, but it is also possible it might only be compatible with 2Mbps rates.

11a I think is compatible with 2Mbps and greater, but not sure how fast.

11n and 11ac should be hunky dory with any rate you so desire.

The effect is pretty mild unless you have either

A) A lot of SSIDs
or
B) A lot of APs
or
C) A very short beacon invertval

Using the standard 1Mbps beacon, at 10 per second, you use roughly 2.5% of your air time, per AP and SSID. It is somewhere in that ballpark.

So if you had 3 SSIDs on the same channel being broadcast by 3 Access Points, you'd be burning close to a quarter of your possibly bandwidth on beacons.

This is part of why living in an apartment complex with a boat ton of routers and SSIDs (and worse when people are running Guess SSIDs on top of a regular one) you can have significant slow down even if it is 4am and no one is actively using their wifi. Far away beacons are going to cause little impact as your client is not trying to listen to them (not while actively receiving/transmiting, except every once in awhile), but closer ones can still cause interference.

So in environments where you need lots of APs and/or SSIDs and/or need frequent beaconing, then turning up the beaconing speed can greatly cut down on how much airtime is being sucked up by the game of Marco Polo.