What's new

Is There a "Penalty" for Using Multiple SSIDs?

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Klueless

Very Senior Member
Now I've been learning that if you can see multiple SSIDs it could suggest possible problems (like if the radio channels they use conflict with mine).

But what if they all come from me?

My router (Asus RT-N66R) allows me to use up to eight SSIDs, four at 5Ghz and four at 2.4Ghz. I use (don't laugh) six. What, if any, penalties or extra overhead am I incurring? Are they of any significance?

Thanks any and all!
 
Those are virtual interfaces, overlaid on the one physical radio. It's no different than using a guest network (or 3).

Using 10 clients with a single radio and 1 SSID will perform the same as using 10 client with a single radio and 10 SSIDs. They're all competing for that single radio.
 
Those are virtual interfaces, overlaid on the one physical radio. It's no different than using a guest network (or 3).

Using 10 clients with a single radio and 1 SSID will perform the same as using 10 client with a single radio and 10 SSIDs. They're all competing for that single radio.
Awesome! Thank you so much for taking the time to reply : -)
 
There may be some small penalty for having to manage and broadcast all of those SSIDs (some people here understand the underlying mechanics far better than me) but it's not going to have any real, perceptible performance impact.
 
Those are virtual interfaces, overlaid on the one physical radio. It's no different than using a guest network (or 3).

Using 10 clients with a single radio and 1 SSID will perform the same as using 10 client with a single radio and 10 SSIDs. They're all competing for that single radio.


I think your statements are wrong.

http://www.revolutionwifi.net/revolutionwifi/p/ssid-overhead-calculator.html

The number of ssid's contribute to worse performance for all.

Including any neighbor AP's that happen to be on the same channel.

The simple answer is that WiFi is time shared. The more ssid's, the less time each has.
 
There's no assumption there as to whether or not "SSID" and "AP" are being utilized synonymously. That's an important assumption.

If each SSID is a separate physical instance, then you have congestion and therefore packet scheduling comes into play - CSMA/CD. If you have multiple physical instances utilizing the same frequency spectrum, it gets increasingly congested and therefore the overhead increases.

However, if only one channel is in use and all SSIDs belong to the same physical interface, you have CPU overhead but the client packet scheduling isn't any different that if you had 1 SSID and 1 physical interface. CSMA/CD is happening at the interface level, not the sub-interface level, at least the way I understand it. It wouldn't be any different than frame processing in a VLAN enabled switch in that sense. Having 10 VLANs on a physical port don't increase Layer 2 congestion, although it increases resource usage on the processor.

This can easily be tested in the real world - enable Guest Network on just about any consumer router available and measure the throughput.

I have two guest networks in my environment and performance on the Private SSID is identical whether they are on or off.
 
Thank you! The chart and text are compelling.

I decided to err on the side of caution and reduced my SSIDs (until I get a chance to benchmark the differences).

Not much I can do about reducing other people's APs other than being attentive and optimizing channel selections I guess.

Thanks again.
 
Last edited:
There's no assumption there as to whether or not "SSID" and "AP" are being utilized synonymously. That's an important assumption.

If each SSID is a separate physical instance, then you have congestion and therefore packet scheduling comes into play - CSMA/CD. If you have multiple physical instances utilizing the same frequency spectrum, it gets increasingly congested and therefore the overhead increases.

However, if only one channel is in use and all SSIDs belong to the same physical interface, you have CPU overhead but the client packet scheduling isn't any different that if you had 1 SSID and 1 physical interface. CSMA/CD is happening at the interface level, not the sub-interface level, at least the way I understand it. It wouldn't be any different than frame processing in a VLAN enabled switch in that sense. Having 10 VLANs on a physical port don't increase Layer 2 congestion, although it increases resource usage on the processor.

This can easily be tested in the real world - enable Guest Network on just about any consumer router available and measure the throughput.

I have two guest networks in my environment and performance on the Private SSID is identical whether they are on or off.


I don't see how it makes much difference? The link provided shows how the overhead is calculated too.

Physical or logical, each ssid has to have it's share of time. And the more time slices used for unique ssid's, the higher the overhead.
 
Each "SSID" in the test is a physical radio competing for airtime on the same frequency. You have the overhead and complexity of two layers of competition, whereas with one radio you only have one.

Like I said, it can be proven in the real world. The OP can run a single SSID and test performance. Then enable all 8 and test it again.

I have had as many as 5 in my environment and right now have 3. If there's any performance detriment, it's certainly not visible to me in doing file transfer or running an app like LAN Speed Test.
 
Each "SSID" in the test is a physical radio competing for airtime on the same frequency. You have the overhead and complexity of two layers of competition, whereas with one radio you only have one.

Like I said, it can be proven in the real world. The OP can run a single SSID and test performance. Then enable all 8 and test it again.

I have had as many as 5 in my environment and right now have 3. If there's any performance detriment, it's certainly not visible to me in doing file transfer or running an app like LAN Speed Test.


One radio with multiple ssid's or unique ssid's with multiple radio's the time slice needed is the same, regardless of what additional overhead the processor sees.

'In your environment' is not a general statement, it is specific. Not only for nearby AP's (or lack of them, on the same channel) but also the router and firmware and drivers used. Currently with the 3 ssid's you fall into the 'Low' overhead as categorized by that link, so one more reason you won't see it.

A WiFi time slice is not variable, that is why multiple ssid's create additional overhead. A single radio cannot magically send beacons of multiple ssid's at the same time and neither can multiple radios send their beacons without waiting their turn from the others in the immediate area (whether same ssid or not).
 
Well, I did a little research using some Aruba resources and it appears that the answer to this question isn't a simple yes or no.

The real answer is that multiple VAPs absolutely CAN affect performance depending on the number of VAPs, the number of clients, and the band being used.

Obviously total bandwidth available is limited, especially on 2.4Ghz. The performance penalty comes not from contention for AP time resources as much as it comes from management traffic generated by each VAP and the associated clients.

A 2.4Ghz AP with 3 VAPs and 100 "seen" clients can use almost 40% of the total available bandwidth (~22 of the 54Mbps) just for management traffic. Obviously, that a significant impact.

That being said, that number drops off with the number of clients. In the majority of home setups, the total bandwidth used by management traffic is going to be less than 5% of total bandwidth available.
 
Well, I did a little research using some Aruba resources and it appears that the answer to this question isn't a simple yes or no.

The real answer is that multiple VAPs absolutely CAN affect performance depending on the number of VAPs, the number of clients, and the band being used.

Obviously total bandwidth available is limited, especially on 2.4Ghz. The performance penalty comes not from contention for AP time resources as much as it comes from management traffic generated by each VAP and the associated clients.

A 2.4Ghz AP with 3 VAPs and 100 "seen" clients can use almost 40% of the total available bandwidth (~22 of the 54Mbps) just for management traffic. Obviously, that a significant impact.

That being said, that number drops off with the number of clients. In the majority of home setups, the total bandwidth used by management traffic is going to be less than 5% of total bandwidth available.


You are only seeing one side of the equation.
 
Interesting discussion. Always wondered how multiple SSID's from a single radio works.

Does each SSID have to have it's own beacon? If so how are the beacons arranged? One SSID per beacon interval?
 
Multiple SSID's do incur overhead for each one - and that's less airtime for the channel...

Only use what one must... if "Guest" networks aren't really needed, then turn them off...
 
Does each SSID have to have it's own beacon? If so how are the beacons arranged? One SSID per beacon interval?

Yes.. and not just beacons, but other management frames as well...
 
Using 10 clients with a single radio and 1 SSID will perform the same as using 10 client with a single radio and 10 SSIDs. They're all competing for that single radio.

Actually, no - 10 clients in one BSS (SSID) vs. 10 BSS's with 1 client each... each SSID/BSS/Beacon/other management frames exacts a toll on airtime and chipset compute bandwidth... each having it's own schedule for CSMA/CA scheme for that BSS...

The overhead will eventually kill the channel - saw this first hand with MESH development...
 
anyone know how decryption/encryption affects performance with multiple SSIDs? Seems like the layer 2.5 firmware has to receive a frame or frames, look at the SSID, lookup the keys du jour for that SSID, en/de-crypt. Perhaps it matters not, as this load is simply the frame/packet rates in aggregate, no matter the number of SSIDs.

However, does the router have to generate a beacon for each SSID or can the SSIDs all be batched in one beacon? If the former, then there could be n times 10 beacons per second? That gets to be a lot of air time (n = number of SSIDs).
 
One of my take-aways from this most enlightening conversation is when I'm dealing with a lot of clutter I can now:
  • Reduce the number of my SSIDs
In addition to:
  • Relocating my AP
  • Changing (optimizing) channels
  • Moving to 5Ghz (if/when possible)
By the way ... many thanks to everyone!
 
Last edited:

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top