Bad Downlink Efficiency? But Good Uplink...

What are some reasons the downlink is suffering compared to the uplink? His RSSI stays around 1750, Jitter 5/6, and Power (-62dmb)

Stats for LUID: 5 Test Duration: 10
Downlink RATE: 325632 bps
Uplink RATE: 291123 bps
Downlink Efficiency: 26 Percent
Max Downlink Index: 97
Actual Downlink Index: 25
Expected Frag Count: 6360
Actual Frag Count: 25037
Uplink Efficiency: 100 Percent
Expected Frag Count: 5686
Actual Frag Count: 5686

Device type 900MHz - Multipoint - Subscriber Modem - 0a-00-3e-90-51-da
Software Version CANOPY 6.1 Dec 08 2004 09:22:02 SM-DES
Software Boot Version CANOPYBOOT 3.0
FPGA Version 111504 (DES)
Uptime 8d, 15:29:04
System Time 22:40:58 01/10/2000
Ethernet Interface 100Base-TX Full Duplex
Subscriber Modem Stats
Session Status REGISTERED
Registered AP 0a-00-3e-90-1f-7c
RSSI 1775 (-62 dBm)
Jitter 6
Air Delay 215 (approximately 5.99 miles (31605 feet))

AP List
Current entry index: 0

Index: 0 Frequency: 919.0 MHz ESN: 0a-00-3e-90-1f-7c
Jitter: 7 RSSI: 1760 Power Level: -63 Beacon Count: 7 BRcvW: 1
Type: Multipoint Avail: 1 Age: 0 Lockout: 0 RegFail 0 Range: 10535 feet TxBER: 0 EBcast: 0
Session Count: 2 NoLUIDS: 0 OutOfRange: 0 AuthFail: 0 EncryptFail: 0 Rescan Req: 0
FrameNumber: 637 SectorID: 0 Color Code: 1 BeaconVersion: 1 SectorUserCount: 28
PercentDataDwn: 75 PercentDataUpHi: 0 NumDAckSlots: 3 NumContSlots: 3 NumUAckSlots: 3

This link does go through a few groups of trees. The customer has not called in complaining, but I am just curious and still learning. Thanks


If you’re running in 2X mode, switch the SM to 1X and run the Link Test again.

Did Version 6.1 have that capability? Running Software Sched. also

Oops! No, 2X first appeared in 7.0.

Because the uplink was good, I didn’t think interferrence was very likely. You should probably start with the Spectrum Analyzer on the SM, but that requires bothering the customer, either on the phone or with a visit.

wirelessSolutions wrote:

... snipped ...

Jitter 6

... snipped ...

Jitter: 7

... snipped ...
This link does go through a few groups of trees. The customer has not called in complaining, but I am just curious and still learning. Thanks


Too much noise on the link. Interesting that the CPE seems to be more sensitive to it than the AP, but 900 MHz hardware doesn't like jitter's above 4, even with software scheduling.

I aggree with Micers. you’ve got a huge fade margin, but that Jitter would be a show stopper for me.

Could you clear some of those trees with more height on the Customer end?

Something else that was giving me some problems at one site was the radio being rotated abit. The polizaration was off, AP was level, but the SM was “leaning” several degrees. Fixing that dropped a 3-4 jitter to 1-2.

Hope that helps,

This SM is mounted on a million +$$$ house. I don’t think we can get it any higher. We could definately try a yagi. But since we havnet heard the customer complaining, we’ll leave it be. We have many links that hover around 4-5 jitter… I CANT WAIT TIL FALL!



I can see your point, definately. Consider this though, how much “flapping” will your customer tolerate (or grumble to his friends/family/neighbors) before calling to complain.

If you are providing what you feel to be an unacceptable level of service, then it should be fixed – weather the customer complains or not.

I can understand your wait-and-see attitude. The fact that he has not called, does not necessarily mean he’s completely happy with your service though. Word of mouth is a powerful tool, and it can swing both ways.

I called this customer last week letting him know we recognized lately his link is poor. We are looking into assembling a mast for him, which I hope helps.


If you don’t mind, could you please post how it goes? I’m currious if it makes a diference for you/him.