Power AP-N Poor Throughput?

[ry12a]

Hi
I have a Power AP-N
Firmware Version: XM.v5.3.3
Build Number: 9634

I am only seeing transfer rates of max 2.5MB/Second

I also have some other access points that consistently allow transfers at 4.5 MB/Sec

I have it set up as an Access Point/Bridge as you can see by the attached files.
I am connecting to a gigabit switch and have no other network issues.
To be honest I'm a bit disappointed!
Can anyone shed any light?

Thanks for looking!

 

[stevech]

That's
http://ubnt.com/powerapn
right?

have you done tests to conclude that the client PC is not the cause?

The WAN port goes by cat5 to your router?
Are the LAN ports disabled in bridge mode?

 

[ry12a]

Thanks for looking!
Yes indeed that's the device...and it is connected by cat 5 to the switch!
I am using the same source device and files to achieve the results...
Not sure about disabling the LAN ...I lose connection to the webpage for the device then....

 

[keenanj]

you say you have other networks if so you will have interference you can ony have one 40Mhz 11n device in one area.

This could be your issue.

 

[rhombus]

Try turning the output power down till it allows a faster modulation index of MCS 15 (max link rate 130Mbps without SGI). Or set either or both manually.

Currently you have 28dBm output power, and MCS 12 (max link rate 78 Mbps).

 

[stevech]

some/most 802.11g/n gear that promises more than 50 or 100mW transmit power - and costs less than $150 or so, get the high power via an amplifier that has inadequate "linearity" for the higher speeds of 11g and 11n. The technical reason is that the amplifier can distort the signal and cause a higher bit error rate. This leads to lower net yield throughput at the IP layer despite an apparently higher raw 802.11 bit rate.

A bit more drill-down: 802.11's OFDM signals have a high peak-to-average power ratio. The ratio is larger as the raw bit rate (modulation order) increases. The amplifier needs to be highly linear at the peak power portions of the modulated signal (to simplify). Getting that good linearity at high power adds a great deal of cost. So it is common practice to use "OFDM backoff". This means reduce the transmitter's power by about 5dB for the higher speeds (modulation orders) of 802.11. The bit error rate will reduce causing the net throughput to improve (this is called "goodput").

The amplifier's linearity matters far less in 802.11b mode, and the lowest speeds of 11g/n. So the power can be "max" for these speeds.
Remember too that the speeds are chosen dynamically for each client device, so this is a catch-22 of sorts. Some products allow you to cap the max modulation order (802.11 bit rate) to match the chosen transmitter power. The really good professional gear adjusts the transmit power frame-by-frame to each client, assuming a "balanced" link is desired: that being the AP and the client use the same radiated power.


Yes, the reduced power costs a bit in range, but of course, the range is limited NOT by the access point/router's transmitter power, but by the client (laptop?) transmitter power! So even a megawatt access point won't fix the problem of a weak transmited signal from a laptop.

 

[rhombus]

but of course, the range is limited NOT by the access point/router's transmitter power, but by the client (laptop?) transmitter power! So even a megawatt access point won't fix the problem of a weak transmited signal from a laptop.

Yet, SNB's charts show it is not just possible, but is in fact the trend, that dowlink results are in general significantly better than uplink.

Therefore even to a laptop, high power (if done well, and yes it is a big if) can prove a useful design choice all other things equal, especially to multiple clients or for wider coverage with less directional antennas.

And if you have engineered an AP/router with an extra high receive sensitivity, that would go to waste without extra high transmit power to complement it.

 

[Mangusta1969]

I don't have any throughput problems with my setup. From pingtest.net, I consistently receive the following results from my 6 month old PowerAP-N:

Packet Loss 0%
Ping 34 msec
Jitter 2 msec
Grade A


From speedtest.net, I consistently receive the following results:

Ping 21 msec
Download speed 5.09 Mbps
Upload speed .65 Mbps


I get those numbers about 200 feet away from the AP-N and with the wireless signal going up one floor, penetrating three walls and also going through several mature trees in the yard that separates the PowerAP-N location from my principal client locations. My performance results are also pretty consistent with a Mac Airbook, a Mac Powerbook, and a Mac IMac.

My signal source is a Zyxel DSL modem that is hard-wired to the WAN Port of the PowerAP-N. My service is ATT Elite DSL, which claims 6 Mbps down and .5 Mbps up speeds. I have verified with the same Mac portable clients that were temporarily hard-wired to the PowerAP-N that the above download and upload speeds are being limited by the ATT Elite DSL service and not by the PowerAP-N's wireless speeds. I also tried a direct Mac portable Cat5 connection to the Zyxel modem and did not notice any appreciable speed changes with the PowerAP-N completely out of the system; I don't have those direct connection, all wired, numbers handy right now.

I have 4 other interfering 2.4 Ghz networks in my neighborhood. They used to wreak havoc with a Cisco E4200; I had to reboot that devil several times per week to maintain acceptable and stable wireless throughput speeds. The PowerAP-N deals with those competing networks without any problems. I have a software firewall configured in my Mac computers. I did not try firmware based firewall rules in the PowerAP-N itself. Are you using any firewall setups?

As Stevech suggested, you should try to eliminate variables by trying a different client(s) with your PowerAP-N and report back to us with your throughput and ping/jitter findings. If you see wireless speed and ping/jitter issues that are significantly slower than your wired speeds, you can also use the PowerAP-N's RF spectrum scanning feature to identify whether or not you have signal interference problems with nearby and competing networks. My PowerAP-N just seems to overwhelm the nearby networks with its much stronger transmit signal and 5 dbi gain antennas for both transmit and receive directions; on the receive side, those vertical omni antennas help significantly with weak client signals from laptops and with lots of obstacles between the transmit and receive devices.

Also take a look here, as PowerAP-N users have not reported speed limitations like you are experiencing:

http://forums.smallnetbuilder.com/showthread.php?t=5312