Senao bridge running slow.

Hi.

I've got 3 sites all linked together via Senao SL-3054CB3 PLUS bridges (so that's 4 bridges). the short link is about 500m, using 8dBi Yagi antennas. It runs great, with FTP speeds of around 3MB/s.

The long link is about 5km, using 24dBi parabolic grid antennas. The Senaos report good "communication strength" (83% - about the same as the short link), but the link runs very slowly. FTP gets about 9kB/s. I've tried a whole range of channels, and some are worse than 9kB/s, but none are better.

The parabolic grids are the 2nd pair of antennas that I'm trying for the long link - the 1st pair was some 15dBi Yagis. The Senaos report better "communication strength" with the parabolic grids (up to 83% from about 73% with the Yagis), but both maxed out at 9kB/s FTP transfers.

There seems to be a short burst of higher transfer speed (maybe up to 20kB/s) in the first second or two of the FTP transfer, but then it comes down to about 10kB/s and eventaully settles around 9kB/s.

The Senaos on the long link are set to "802.11b only" - they seemed to run slower on mixed b and g mode, or g mode only.

Does anyone have a clue why the short link is working so well and the long one is so bad, when they both have the same "communication strength" according to the Senao's? Any ideas on how to fix it?

Since I've swapped the antennas already, I'm thinking of trying out a pair of Netgear bridges instead of the Senaos. Has anyone else got a long range link (around 5km or more) running at a good speed? If so, what equipment are you using?

Thanks for any help.

Woofda

Is there constant traffic on the links?
Remember that 802.11 is half-duplex and that you may also suffer from hidden node problems

Perhaps a little ascii diagram of your setup will help a bit.

Also, are you experiencing packet loss on the long link?
Some ping statistics?

No idea about this, but thumbsuck... wouldn't ack's for packets receive take longer to return to the sender over a long link? So on a short link it's quick because it doesn't travel that far, but on a longer link it does? 9kb does sound low though...

Okay - here's an ASCII diagram of the setup:

A <---0.5km---> B <-----------5km----------> C

The A<--->B link can ftp at 3MB/s, the B<--->C one at only 9kB/s.

A has an 8dBi yagi, B has an identical 8dBi and a 24dBi parabolic grid (on the same mast, aiming about 35 degrees away from each other, with the parabolic grid about 80cm higher than the yagi), C has a 24dBi parabolic grid. A<--->B is running on channel 4. I've tried B<--->C on a wide range of chanels (2, 6, 7, 8, 9, 12) and the best I've got is 9kB/s ftp.

The Senao bridges report 83% "communication strength" on both the A<--->B link and the B<--->C link. The B<--->C link went up to 83% from 73% when I replaced some 15dBi yagis with a pair of 24dBi parabolic grids.

I'm not sure how much the Senao's reported "communication strength" can be trusted, because it doesn't seem to refresh very often and stays reporting a high signal strength for quite a while when the other side of the link goes down.

Here's the ping on the bad link (B<--->C):

Pinging 192.168.XXX.XXX with 32 bytes of data:

Reply from 192.168.: bytes=32 time=70ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=20ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time<10ms TTL=64
Reply from 192.168.: bytes=32 time<10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time<10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time<10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=10ms TTL=64
Reply from 192.168.: bytes=32 time=20ms TTL=64

Ping statistics for 192.168.XXX.XXX:
Packets: Sent = 17, Received = 17, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = 70ms, Average = 12ms


Bridge config at C:

Access Point Information
State: Associated (ad-hoc mode)
Wireless Network Name (SSID): XXXXXXXX
Channel: 2
Transmission Rate: Best (automatic)
Communications Strength: 84%
BSSID: XXXXXXXXXXXX
WEP: enabled
WPA: disabled

Wireless Network Type: Point to Point
Wireless Network Name (SSID): XXXXX
Desired BSSID: XXXXXX
Channel: 2.4 GHz channel 2
Transmission Rate (Mbits/s): Best (automatic)
Wireless Mode: 802.11b Only

Does that help?

Woofda

In reply to daffy, no, there isn't constant traffic on the long link - none at all except for the FTP that I'm using to test.

at site B, I assume you're using 2 Senao bridges and not a splitter?

Yup, 2 seperate bridges at B (so 4 in total for the whole A,B,C setup.. I've tried swapping the Senaos on the fast link with the ones on the slow link to eliminate the possibility of a fault with one of them. The slow link was still slow, and the fast one still fast.

Try fixing the link speed on B-C to something like 2mbs and do a test again. With auto-rate selection I've seen AP's continually reselecting a rate which slows things down.

Have you tested the same with WEP disabled?

Start with 2mbit on 802.11b, run tests. Remember that 802.11b rates are 1,2,5.5 and 11.
Anything else is 802.11g.

As TheRoDent said, fix the rates, dont let the AP's select Automatically.

Yeah.. WEP will cause a slowdown too.

I've noticed a similar scenario. I'm utilising what seems to be the Senao SL-3054CB3 unit built in with a 12dbi flat Panel all in a water-proof casing with about 20m network cable - power is supplied to the unit via the spare lines in the cat5 cable.

A ----> B
.\ .\
.. \. \
... \ ..C
.... \
..... D


Point A is set up as an Access Point, the other three as bridges. About a 45 degree spread from A between B & D. B about 700 metres. C about 1 Km and D about 4 to 5 Km. Signal strength similar on B, C & D about 78-80% Yet B and C reaches speeds of about 13Mbs where D just under 1Mbs. I will experiment this weekend with different settings to try & improve this. These units come from Miro - www.miro.co.za

hey guys, i know this is a bit off your topic but i can't get anyone else to answer this question, im trying to setup a wireless connections, g or b, i only need speed of about 2mpbs constant, distance needed to cover is about 1km , theres a few trees in the way but its basically over a school field so theres very little in the way, i'm also on a budget here. any ideas of the hardware i should get, i'm thinking of 2 AP's , and both with simple antennas on. but makes and models and what antennas i do not know!

patrick123 said:

Point A is set up as an Access Point, the other three as bridges. About a 45 degree spread from A between B & D. B about 700 metres. C about 1 Km and D about 4 to 5 Km. Signal strength similar on B, C & D about 78-80% Yet B and C reaches speeds of about 13Mbs where D just under 1Mbs. I will experiment this weekend with different settings to try & improve this. These units come from Miro - www.miro.co.za

Click to expand...

Welcome to the 802.11 "hidden node" problem. Your B,C,D bridges are all contending for the same channel at the access point. B,C with stronger signal and less noise will be able to grab the channel more than D will ever can due to the fact that the AP can hear them clearly.

Thanks RoDent, it seems the easiest solution would be to setup either a repeater at point C, or get another unit on another channel between A & D.

wouldn't it just be simpler to set point B in ap mode and let poing A and C (bridge mode) connect to B?? Everything should effectively work better this way!

Unfortunately I cannot let point B be the AP as these units I have are built into a weatherproof casing the complete unit has about a 2inch antenna cable straight into a flat panel antenna. If push comes to shove, I suppose I could open the unit at point B use an Omni instead.

What about using a splitter with two grids ? Better gain than the omni.

No.
No Splitters, end of story.
Antenna to Radio ratio is ALWAYS 1-to-1. No Exceptions

These units are both b&g. I've now forced b only at a limit of 5.5Mbs. I can now get a steady 5Mbs from all three stations, lost on the swings, but made up on the merry-go-round.

Interesting. On 802.11b, 5.5 and 11 use the same rf modulation. Its not very often that you'll see 5.5 work and 11 wont.

Try playing with the lower speed 802.11g settings, like 6,9 and 12.