Router help EDIT: Got Asus RT-N13U

Now if the client only has a link speed of 30mbs to the repeater, it will have a higher throughput than...

If the repeater linked at say 50Mbs, and repeated at 25Mbs.
This is exactly why i said you haven't understood repeating :p

The reason we assume the repeated speed is half is because Wifi is half-duplex : the air-time it spends to receive cannot simultaneously be used to transmit. A client-repeater link at 30Mbps, can get an assumed 1/3rd throughput of 10Mbps ONLY if the repeater is transmitting data for 100% of the air-time. But its not since it spends a portion of the air-time receiving data from router 1.

Here's an image of what I'm saying (sorry about the cheesy drawings :p) :
bHiQv6A.gif


------------------
Here's what I proposed for optimal speeds and why "repeating" speeds are halved in the first place. 75Mbps receive using 50% of the air-time/75Mbps transmit using other 50% = net speed of 37.5Mbps :
AklLY3h.gif
 
[Off-topic]

Additionally, these are all n clients, if there were any g or b clients in here they would be bottlenecks regardless of how close they were to the router. As the radio can only talk one protocol at a time. Slower speeds with b or g mean slower speeds for everybody else for the duration of the conversation. So to prevent this the router is setup as 'n only' clients and not mixed mode.
This is another concept you have misunderstood. The router broadcasts to all clients but the transmission is intended for only one client at a time. The n clients aren't talking at b or g speeds - the b or g clients are using proportionately more air-time which leaves less air-time for the n clients. Understanding air-time is what will help you understand the repeater concept too.

[/Off-topic]
 
nooo let it be on mine! hahahah i am tryin to understand as much as i can and its already subscribed! hahaha :p

ok but how do i set it up for a optimal speed setup, like the half one u've shown?
 
ok but how do i set it up for a optimal speed setup, like the half one u've shown?
Simple, there are two competing proposals here, try both out and let us know the difference.

My proposal:
3 walls at 40 feet

router (1)|.............. (repeater)(2)|................. (3) | client

1 wall between router-repeater, 2 walls between repeater-client

Eddy's proposal:
router (1)|.................................(2)|(repeater)................. (3) | client

2 walls between router-repeater, 1 walls between repeater-client

or to put it another way, Eddy wants you to position the repeater in such a way that link speeds router-repeater is as close to same as repeater-client.

I think its possible to squeeze more throughput out by giving little more link speed to router-repeater (20-40%) more say than repeater-client.

Now how to measure the link speeds here ?

Have to examine the status page reported by the repeater to see what link speed it got from the router. Client link speed is easy to see, just hover over the wifi icon in the tray on the dell.

You will have to position the repeater in various locations, might need to use an extension cord for power. And then on the client copy files over from your room and see what the throughput on the dell is.

Course before you are able to do this you will have to setup the repeater properly to begin with, can be a bit quirky from what i've read. I read a comment from someone who got it working here and it stayed up for over a month. No need for constant reboots.
 
[Off-topic]


This is another concept you have misunderstood. The router broadcasts to all clients but the transmission is intended for only one client at a time. The n clients aren't talking at b or g speeds - the b or g clients are using proportionately more air-time which leaves less air-time for the n clients. Understanding air-time is what will help you understand the repeater concept too.

[/Off-topic]
I've understood the airtime concept so lets consider a scenario with n-only clients. To keep it simple no repeater involved here.

Lets say you have a n client at the edge of just one routers range. Are you saying it will slow down the speeds of faster n clients that are closer ? Because it hogs airtime disporportionately in comparison to the others.

Basically it should only take one slow, busy, n client to bring the mighty 'dark knight' to its knees.

Just one N150 mobile/tablet at the edge of the routers range slows down everything for everybody. Taking up very little spectrum because its a weak connection yet its as if most of the spectrum were occupied.
 
ok but how do i set it up for a optimal speed setup, like the half one u've shown?
I don't have the N13u, so hopefully it reports link speeds in the status page as blr_p says. Windows usually exaggerates client link speeds, but I've seen people say that Intel cards/drivers tend to report more accurate speeds.


I think its possible to squeeze more throughput out by giving little more link speed to router-repeater (20-40%) more say than repeater-client.
Data in data out. Where will the excess throughput the repeater gets from the higher link speed go?

If you're still sticking to 'client link is 30Mbps so its faster', you still haven't understood airtime. Lets try numbers:-
30Mbps = 30 Mega bits per second. Take a sample transmission length of 1 second.
Lets say the repeater radio is busy for 0.25 seconds while its receiving data from the first router; the client radio will be idle for these 0.25 seconds. That leaves 0.75 seconds for the repeater radio to communicate with the client. 30Mbps x 0.75s = ~22.75Mb.


Basically it should only take one slow, busy, n client to bring the mighty 'dark knight' to its knees.
Probably. Airtime is shared, it depends on who gets it. A slow busy link uses more airtime, which means it has a higher chance of blocking the channel. Also AFAIK there are no standards in the wi-fi spec on how to prioritise traffic, so its upto router manufacturers on how they handle this.

Edit: "Airtime fairness" is the buzzword manufacturers are using to tackle what I'm describing. Here's an article about it : http://www.cwnp.com/cwnp_wifi_blog/what-is-fairness-anyway/ . And quoting : "The problem with PHY-based fairnes is that an 802.11n client may be at a cell's edge transmitting at 12 Mbps while an 802.11a client may be close to the AP transmitting at 54 Mbps. PHY-based fairness is only fair depending on positioning of clients (and thus the data rate of those clients at any point in time)."

Taking up very little spectrum because its a weak connection yet its as if most of the spectrum were occupied.
Wrong. Read up on spectrum first.
Spectrum width doesn't vary based on weak or strong signals, it's always 20/40MHz. And all clients - weak or strong signals - will use the same centre frequency, i.e. channel. If one client occupies the channel, nobody else can use it till its free.
 
I don't have the N13u, so hopefully it reports link speeds in the status page as blr_p says. Windows usually exaggerates client link speeds, but I've seen people say that Intel cards/drivers tend to report more accurate speeds.
Checking the screens & manual and the N13 does not indicate what link speed it got from the router :|

All you see is little icons with bars for strength for various networks available.

Throughput test on the client end is the only way.
 
How to test throughput between nodes in a router <> repeater setup ? I have a WDS setup and i can see the routers connect at 270 Mbps, but how to see what is the actual throughput that i get across this setup ?
 
How to test throughput between nodes in a router <> repeater setup ? I have a WDS setup and i can see the routers connect at 270 Mbps, but how to see what is the actual throughput that i get across this setup ?
I was thinking of starting a thread about this but the quick & dirty way is file transfer.

Either with FTP (non-windows clients) or via network shares.
 
sister finally has wifi in her room, thanks alot for the help everyone! i got the setup up and running via the asus utility / disabled the beetel wifi / and everything seems fine for now.. will test more in a few days. anything i should be checking?
 
Transfer some files across and note the transfer rate. Use files that are a few hundred MB's or more.

You saw the little diagram i put up earlier, with the walls, tell us which one you went with. Then try the other and note the transfer rate. Want to know which one works better.

Thing with your setup is to see whether you can go more than a couple of days without having to reset the repeater.
 
sister finally has wifi in her room, thanks alot for the help everyone! i got the setup up and running via the asus utility / disabled the beetel wifi / and everything seems fine for now.. will test more in a few days. anything i should be checking?
Still waiting to hear on what the transfer rates are between the main router & the repeater.

Given you've not posted since i take it the setup is stable.
 
bro i didnt check it only. one of the routers died right now..i had alot of issues setting it up.. my xbox acted funny and kept dc'ing all the time.. yesterday i called a friend and did the proper setup.. might have a faulty cable.. and the other router died aswell.. so now waiting for it to come back... i hadto keep restarting it while keeping it furthest, now ill plan to keep it in the middle and setup once i get em back. once i do.. i will pm and find out how to test transfer speeds :(
 
oh, why didn't you post back here. Since you said your sister has wifi one would have thougt you figured out everything and got it working.
 
bro i didnt check it only. one of the routers died right now..i had alot of issues setting it up.. my xbox acted funny and kept dc'ing all the time.. yesterday i called a friend and did the proper setup.. might have a faulty cable.. and the other router died aswell.. so now waiting for it to come back... i hadto keep restarting it while keeping it furthest, now ill plan to keep it in the middle and setup once i get em back. once i do.. i will pm and find out how to test transfer speeds :(

Updates ?
 
Back
Top